{"family":"memory","name":"Agent Memory","count":311,"license":"CC-BY-4.0","systems":[{"id":"0-5-physical-intelligence--pi-website","name":"π0.5 (Physical Intelligence)","url":"https://www.pi.website/blog/pi05","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Multi-Scale Embodied Memory (short + long episodic)","license":"Apache-2.0 (openpi repo; repo also ships a LICENSE_GEMMA.txt for Gemma-derived terms)","latest_release":"","deployment":"Cloud inference + on-robot deployment (websocket streaming)","pricing":"searched not found","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"Physical-Intelligence/openpi — ~13.3k stars, last commit not confirmed (π0.5 base, DROID and LIBERO checkpoints released)","storage":"vector","retrieval":"attention|similarity","persistence":"long-term","update":"append-only|consolidation","unit":"episode|trajectory","governance":"opaque","conflict":"append","description":"Extends π0 with MEM layer giving the policy short-term (within-task) + long-term (cross-task, >10-min horizon) memory. Co-trained on robot teleoperation + human video + text. Hierarchical inference: VLA predicts subtasks as text tokens, then executes as action chunks.","cited_cells":33,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/0-5-physical-intelligence--pi-website"},{"id":"abridge--abridge-com","name":"Abridge","url":"https://www.abridge.com/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Grounded-transcript provenance","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"file|vector","retrieval":"similarity|extraction-pull","persistence":"long-term","update":"append-only","unit":"episode|document","governance":"auditable","conflict":"editor-in-the-loop","description":"Clinician-assist ambient documentation. Source mapping: every AI-generated summary element traced back to the source utterance. Audit-and-trust layer over episodic memory. Built on proprietary 1.5M+ medical-encounter dataset.","cited_cells":59,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/abridge--abridge-com"},{"id":"acontext-memodb--gh-memodb-io-acontext","name":"Acontext (memodb)","url":"https://github.com/memodb-io/Acontext","section":"Dedicated memory layers","tier":3,"status":"archived","last_verified_at":"2026-10-29","type":"Framework","license":"Apache-2.0","latest_release":"last commit 2026-04-21","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/memodb-io/Acontext (3696★)","storage":"file|relational","retrieval":"agentic|exact-match","persistence":"long-term","update":"extraction|agent-controlled","unit":"skill|file","governance":"editable","conflict":"llm-arbitrate","description":"'Agent Skills as a Memory Layer': an open-source skill-memory layer that stores what agents learn as agent skill files. Built by the Memobase team. The README now says Acontext is no longer maintained; the team has moved on to alook.ai.","cited_cells":12,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/acontext-memodb--gh-memodb-io-acontext"},{"id":"acuvity-now-proofpoint--acuvity-ai","name":"Acuvity (now Proofpoint)","url":"https://acuvity.ai/ai-runtime-security/","section":"Memory governance, privacy & safety","tier":1,"status":"acquired","last_verified_at":"2026-10-23","type":"MCP / Shadow-AI runtime enforcement","license":"","latest_release":"","deployment":"on-prem or cloud (Proofpoint enterprise deployment); agentless approach","pricing":"enterprise pricing via Proofpoint (not public); contact sales","mcp_support":"Yes — Secure MCP servers (Model Context Protocol) is a core capability; protects MCP-based agent integrations","a2a_support":"no data — searched proofpoint.com/us/platform/ai-security, acuvity.ai/, acuvity.ai/integrations/, acuvity.ai/blog/; A2A protocol not advertised","backend_storage":"custom","llm_lock":"","github":"","storage":"n/a","retrieval":"agentic","persistence":"session","update":"read-only","unit":"prompt","governance":"auditable","conflict":"n/a","description":"Acquired by Proofpoint (closed 2026-02-12). Runtime enforcement targeting memory poisoning, unauthorised execution, identity spoofing per the OWASP LLM threat list. Visibility/control over MCP servers and locally-installed AI tools — the infrastructure layer where memory is most exposed.","cited_cells":39,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/acuvity-now-proofpoint--acuvity-ai"},{"id":"adaptive-rag--gh-starsuzi-adaptive-rag","name":"Adaptive-RAG","url":"https://github.com/starsuzi/Adaptive-RAG","section":"Retrieval-as-memory hybrids","tier":3,"status":"research-code","last_verified_at":"2026-10-23","type":"Query-complexity routing","license":"Apache-2.0","latest_release":"no releases","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"starsuzi/Adaptive-RAG — ~412 stars, last commit unknown (repo page did not show date)","storage":"vector","retrieval":"similarity|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Smaller classifier LM predicts query complexity, then routes to no-retrieval / single-step / iterative retrieval as appropriate. NAACL 2024.","cited_cells":27,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/adaptive-rag--gh-starsuzi-adaptive-rag"},{"id":"agentmemory-rohitg00--agent-memory-dev","name":"agentmemory (rohitg00)","url":"https://www.agent-memory.dev/","section":"Coding-agent memory","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"Apache-2.0","latest_release":"last commit 2026-10-21","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/rohitg00/agentmemory (29786★)","storage":"hybrid|kv|vector","retrieval":"hybrid-rerank|injection","persistence":"long-term","update":"consolidation|extraction|evict-oldest","unit":"episode|summary|fact","governance":"user-controllable|inspectable","conflict":"last-write-wins","description":"Self-hosted persistent memory layer that captures coding-agent activity and recalls it in later sessions, positioned on public benchmark results. #7 in TeleAI Awesome-Agent-Memory","cited_cells":20,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agentmemory-rohitg00--agent-memory-dev"},{"id":"agentops--agentops-ai","name":"AgentOps","url":"https://www.agentops.ai/","section":"Memory observability & monitoring","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Operational metrics for agent memory","license":"MIT (SDK)","latest_release":"v0.4.21 (2025-08-29)","deployment":"SaaS cloud; SDK integration into agent code","pricing":"free tier; Pro/Enterprise pricing not publicly listed; contact sales","mcp_support":"via official adapter","a2a_support":"no data — searched agentops.ai, docs.agentops.ai, docs.agentops.ai/v2/integrations, README on AgentOps-AI/agentops; A2A protocol not advertised (CrewAI/AG2/LangGraph/Autogen integrations only)","backend_storage":"custom","llm_lock":"","github":"AgentOps-AI/agentops — ~5.8k stars","storage":"relational","retrieval":"exact-match","persistence":"long-term","update":"append-only","unit":"episode","governance":"auditable","conflict":"n/a","description":"When Mem0 is connected, gains Memory Operation Timeline, Search Analytics, Memory Growth tracking, Error Tracking per memory call. Standalone, records context at each step but doesn't analyse memory quality.","cited_cells":47,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agentops--agentops-ai"},{"id":"agents-md--agents-md","name":"AGENTS.md","url":"https://agents.md/","section":"File-backed / editor paradigms","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Cross-tool open standard","license":"MIT","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"file (AGENTS.md)","llm_lock":"","github":"agentsmd/agents.md — ~25k stars","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"manual","description":"Open Markdown 'README for agents' at repo root; nested files in subprojects take precedence (closest wins). Holds build/test commands, conventions and PR rules. Stewarded by the Agentic AI Foundation (Linux Foundation); supported by Codex, Jules, Cursor, Copilot, Gemini CLI, Devin/Windsurf and others.","cited_cells":51,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agents-md--agents-md"},{"id":"agixt-adaptive-memory--gh-josh-xt-agixt","name":"AGiXT Adaptive Memory","url":"https://github.com/Josh-XT/AGiXT","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Adaptive provider-spanning + plugin storage","license":"MIT","latest_release":"v1.9.4 2026-04-08","deployment":"Both (OSS self-hosted; agixt.com managed cloud implied)","pricing":"OSS free (MIT); agixt.com offers hosted service but pricing not publicly listed","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"multi-vendor (20+ LLM providers)","github":"Josh-XT/AGiXT — ~3.2k stars, last commit 2026-05","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"extraction|overwrite","unit":"episode|fact","governance":"inspectable","conflict":"llm-arbitrate","description":"Open-source AI automation platform. Routes between short-term and long-term memory adaptively across any LLM provider; plugin system for storage backends. Memory managed at the instruction-management layer — task context, instruction state, conversation history as unified agent state.","cited_cells":48,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agixt-adaptive-memory--gh-josh-xt-agixt"},{"id":"agno-phidata-memory--docs-phidata-com","name":"Agno (Phidata) Memory","url":"https://docs.phidata.com/agents/memory","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"AgentStorage + AgentMemory + DBs","license":"Apache-2.0","latest_release":"v3.1.8 (2026-10-21)","deployment":"Both (self-hosted open source; Agno Cloud managed)","pricing":"OSS free (Apache-2.0); Agno Cloud: usage-based, free tier; no per-event fees; no egress costs","mcp_support":"via official adapter — agno MCP toolkit","a2a_support":"searched not found","backend_storage":"pluggable","llm_lock":"any (BYO)","github":"agno-agi/agno — ~42.8k stars, last release 2026-10","storage":"relational|vector","retrieval":"similarity|injection","persistence":"cross-session","update":"overwrite|append-only","unit":"episode|fact","governance":"inspectable","conflict":"overwrite","description":"Agno (formerly Phidata). AgentStorage persists sessions to a DB; AgentMemory auto-classifies/store user preferences and conversation summaries. Single-line integrations with LanceDB, Pinecone, Weaviate, Qdrant.","cited_cells":60,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agno-phidata-memory--docs-phidata-com"},{"id":"ai-singapore-sea-lion--sea-lion-ai","name":"AI Singapore SEA-LION","url":"https://sea-lion.ai/models/","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"SE-Asian cross-lingual semantic-retrieval memory","license":"","latest_release":"","deployment":"Both (free download self-hosted + Amazon SageMaker JumpStart for managed deployment)","pricing":"Free download (Hugging Face + AISG website); no commercial pricing","mcp_support":"not supported","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity|hybrid-rerank","persistence":"parametric-permanent","update":"read-only","unit":"chunk","governance":"opaque","conflict":"n/a","description":"SEA-LION-Embedding (March 2026): retrieval + reranking models contrastively trained on 245M text pairs across 10 SE Asian languages. SEA-BED benchmark (169 datasets). SEA-LION v4 (Gemma-based) at 128K context with native function calling.","cited_cells":42,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/ai-singapore-sea-lion--sea-lion-ai"},{"id":"aider--aider-chat","name":"Aider","url":"https://aider.chat","section":"Coding-agent memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Per-repo chat history file + YAML config","license":"Apache-2.0","latest_release":"v0.86.2 (2026-02-12)","deployment":"Self-hostable","pricing":"OSS","mcp_support":"no — uses CONVENTIONS.md instead","a2a_support":"not supported","backend_storage":"file (.aider.chat.history.md)","llm_lock":"","github":"Aider-AI/aider — ~41k stars (ecosyste.ms) to ~49.7k (releasealert.dev), last commit 2026-02","storage":"file","retrieval":"injection","persistence":"cross-session","update":"append-only","unit":"episode","governance":"user-controllable","conflict":"none","description":"Appends conversation turns to .aider.chat.history.md in project root. restore_chat_history defaults to false — no automatic replay into new sessions. Community /session save command for JSON state snapshots. .aider.conf.yml carries static preferences.","cited_cells":50,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/aider--aider-chat"},{"id":"aider-conventions-md-aider-conf-yml--aider-chat","name":"Aider CONVENTIONS.md / .aider.conf.yml","url":"https://aider.chat/docs/usage/conventions.html","section":"File-backed / editor paradigms","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Multi-file convention pattern","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"file","llm_lock":"","github":"","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"manual","description":"CONVENTIONS.md (any name, loaded via --read or read: in config); .aider.conf.yml (home / git-root / cwd, last wins); .aiderignore (gitignore-syntax).","cited_cells":28,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/aider-conventions-md-aider-conf-yml--aider-chat"},{"id":"aleph-alpha-phariaai-phariacatch--aleph-alpha-com","name":"Aleph Alpha PhariaAI / PhariaCatch","url":"https://docs.aleph-alpha.com/phariaai-home/latest/index.html","section":"Dedicated memory layers","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Sovereign governed-document memory","license":"","latest_release":"","deployment":"Managed cloud (STACKIT EU) or on-prem enterprise deployment","pricing":"Enterprise only (PhariaAI is an enterprise stack; no public pricing)","mcp_support":"not documented publicly","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|file","retrieval":"similarity|extraction-pull","persistence":"long-term","update":"overwrite","unit":"document","governance":"auditable","conflict":"llm-arbitrate","description":"German enterprise stack. PhariaCatch captures + preserves organisational knowledge through structured human-machine collaboration; PhariaAssistant queries via semantic search. Combines sovereign deployment constraints with explainability tooling. The aleph-alpha.com marketing site no longer uses the Pharia names, but the PhariaAI docs still cover PhariaCatch and list releases through v1.260900.0 (Sep 2026).","cited_cells":32,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/aleph-alpha-phariaai-phariacatch--aleph-alpha-com"},{"id":"amplenote--amplenote-com","name":"Amplenote","url":"https://www.amplenote.com/","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Tags + bidirectional + Task Score","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free; Basic $6.99/mo ($71.88/yr); Pro $11.99/mo ($119.88/yr); Founder $29.99/mo ($299.88/yr)","mcp_support":"yes — local MCP server hosted by Amplenote Desktop (127.0.0.1, bearer token) for Claude Desktop/Code and Codex; Unlimited/Founder plans","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph|relational","retrieval":"graph-traversal","persistence":"long-term","update":"overwrite","unit":"document|fact","governance":"user-controllable","conflict":"overwrite","description":"Tags + bidirectional links + task/calendar integration. \"Task Score\" priority matrix. AmpleAI plugin supports OpenAI, Anthropic, Gemini, Grok, DeepSeek, Ollama.","cited_cells":30,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/amplenote--amplenote-com"},{"id":"anda-hippocampus--gh-ldclabs-anda-hippocampus","name":"Anda Hippocampus (now Anda Brain)","url":"https://github.com/ldclabs/anda-brain","section":"Dedicated memory layers","tier":2,"status":"dead","last_verified_at":"2026-10-29","type":"Bio-inspired sleep consolidation","license":"Apache-2.0","latest_release":"v0.13.5 (2026-10-17)","deployment":"Self-hosted on-premises (full data privacy; no cloud service)","pricing":"OSS (Apache-2.0 + MIT); free — no commercial pricing","mcp_support":"yes — built-in MCP server (Streamable HTTP at /mcp/{space_id} and stdio), memory/recall tools","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"ldclabs/anda-brain (renamed from anda-hippocampus) — ~80 stars, last release 2026-10","storage":"graph","retrieval":"graph-traversal","persistence":"long-term","update":"consolidation","unit":"episode|fact","governance":"opaque","conflict":"none","description":"Renamed Anda Brain (repo now ldclabs/anda-brain). Native graph-based memory for autonomous AI agents. Bio-inspired sleep-based memory consolidation. The hosted service brain.anda.ai has been discontinued; Anda Brain is now self-hosted only (Rust service or Cloudflare Worker).","cited_cells":47,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/anda-hippocampus--gh-ldclabs-anda-hippocampus"},{"id":"anthropic-claude-memory--platform-claude-com","name":"Anthropic Claude Memory","url":"https://platform.claude.com/docs/en/agents-and-tools/tool-use/memory-tool","section":"Platform-provider memory","tier":1,"status":"active","last_verified_at":"2026-10-23","type":"File-backed + tool-driven + Auto Dream","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free + paid","mcp_support":"native (first-party) — Claude apps consume MCP","a2a_support":"not supported","backend_storage":"custom (Anthropic-managed)","llm_lock":"","github":"","storage":"file","retrieval":"extraction-pull|injection","persistence":"long-term","update":"consolidation|agent-controlled","unit":"fact|summary","governance":"user-controllable","conflict":"manual","description":"Memory tool API + consumer \"Memory\" feature + Auto Dream consolidation. Document-as-memory rather than extracted-fact-as-memory.","cited_cells":72,"third_party_cells":11,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/anthropic-claude-memory--platform-claude-com"},{"id":"anthropic-managed-agents-memory--platform-claude-com","name":"Anthropic Managed Agents — Memory","url":"https://platform.claude.com/docs/en/managed-agents/overview","section":"Platform-provider memory","tier":3,"status":"active","last_verified_at":"2026-10-28","type":"Filesystem-backed cross-session persistent memory","license":"","latest_release":"","deployment":"Managed-only (hosted Anthropic infrastructure)","pricing":"Claude API token rates + $0.08 per session-hour for Managed Agents","mcp_support":"native (first-party) — Anthropic created MCP","a2a_support":"not supported","backend_storage":"","llm_lock":"","github":"","storage":"file","retrieval":"extraction-pull","persistence":"cross-session","update":"agent-controlled","unit":"file","governance":"inspectable","conflict":"none","description":"Public-beta memory stores for Claude Managed Agents: workspace-scoped collections of text documents mounted into the session sandbox as a directory the agent reads and writes with its file tools; every change is versioned for audit and point-in-time restore, and memories can also be managed via API/Console.","cited_cells":40,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/anthropic-managed-agents-memory--platform-claude-com"},{"id":"apple-notes-apple-intelligence--apple-com","name":"Apple Notes + Apple Intelligence","url":"https://www.apple.com/apple-intelligence/","section":"Personal AI / PKM / lifelogging memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Folder + tag hierarchy + Writing Tools","license":"","latest_release":"","deployment":"On-device models + Private Cloud Compute","pricing":"Closed","mcp_support":"not supported","a2a_support":"not supported","backend_storage":"custom (Apple-managed; on-device + iCloud)","llm_lock":"","github":"","storage":"file|relational","retrieval":"similarity|injection","persistence":"long-term","update":"overwrite|append-only","unit":"document","governance":"user-controllable","conflict":"overwrite","description":"Folder + tag hierarchy; iCloud sync. Writing Tools (proofread / rewrite / summarise) across all note content; audio transcription + AI summary; smart search; reminders from to-dos.","cited_cells":41,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/apple-notes-apple-intelligence--apple-com"},{"id":"asapp-generativeagent--asapp-com","name":"ASAPP GenerativeAgent","url":"https://www.asapp.com/blog/from-models-to-memory-the-next-big-leap-in-ai-agents-in-customer-experience","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Long-term interaction memory + preference","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity|injection","persistence":"long-term","update":"extraction","unit":"episode|fact","governance":"inspectable","conflict":"pii","description":"Treats memory as first-class architecture. Captures the digital footprint of every interaction; retrieves preference and history at engagement time. Public example: airline knowing a frequent flyer wants aisle seats with her son — preference-aware, not just history-lookup.","cited_cells":31,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/asapp-generativeagent--asapp-com"},{"id":"atlas--arxiv-2208-03299","name":"Atlas","url":"https://arxiv.org/abs/2208.03299","section":"Retrieval-as-memory hybrids","tier":3,"status":"research-code","last_verified_at":"2026-10-29","type":"Few-shot retrieval-augmented LM","license":"CC-BY-NC 4.0","latest_release":"n/a","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"facebookresearch/atlas — ~560 stars, unmaintained","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"parametric-edit","unit":"document","governance":"inspectable","conflict":"n/a","description":"Meta AI. Jointly fine-tunes retriever + LM for few-shot tasks. The facebookresearch/atlas repo is no longer maintained; the code is provided as is.","cited_cells":31,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/atlas--arxiv-2208-03299"},{"id":"augment-code--augmentcode-com","name":"Augment Code","url":"https://www.augmentcode.com","section":"Coding-agent memory","tier":1,"status":"acquired","last_verified_at":"2026-10-29","type":"Real-time semantic Context Engine + curated Agent Memory","license":"","latest_release":"","deployment":"Both","pricing":"Standard $20/mo (incl. $20 usage, up to 50 seats); Business $100/mo (incl. $100 usage); Enterprise custom","mcp_support":"yes — Augment Agent is an MCP client (Easy MCP, settings panel, JSON import)","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"extraction","unit":"document|fact","governance":"opaque","conflict":"none","description":"Real-time knowledge graph indexed from connected repos + commit history + CI/CD + tickets + docs. Separate Agent Memory layer (Sept 2025) captures session decisions/corrections/goals as discrete items with developer review/approval UI. MCP endpoint exposes the graph. Acquired by Harness (2026-10-08): Harness bought select Augment Code assets, including Cosmos, the Auggie CLI and the Context Engine, and the team joined Harness. augmentcode.com now leads with Cosmos, a software-factory agent platform.","cited_cells":58,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/augment-code--augmentcode-com"},{"id":"autogen-memory--microsoft-github-io","name":"AutoGen Memory","url":"https://microsoft.github.io/autogen/","section":"Framework-embedded memory","tier":2,"status":"maintenance","last_verified_at":"2026-10-23","type":"List + teachable vectors","license":"MIT","latest_release":"python-v0.7.5 (2025-09-30) — unchanged; no newer release","deployment":"Both (OSS self-hosted Python; Azure AI Foundry managed runtime with GA Q1 2026)","pricing":"OSS (MIT); no direct cost for library","mcp_support":"via official adapter — autogen-ext-mcp","a2a_support":"searched not found","backend_storage":"pluggable","llm_lock":"any (BYO via OpenAI-compatible adapters)","github":"microsoft/autogen — ~60.4k stars, maintenance mode","storage":"kv|vector","retrieval":"injection|similarity","persistence":"cross-session","update":"append-only|extraction","unit":"episode|fact","governance":"inspectable","conflict":"append","description":"AutoGen's memory layer: ListMemory chronological context plus teachable agents that vectorise corrections; integrates with Mem0/Zep rather than building deep memory natively. Microsoft AutoGen is in maintenance mode (community managed, bug and security fixes only), and Microsoft points new projects to Microsoft Agent Framework.","cited_cells":61,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/autogen-memory--microsoft-github-io"},{"id":"autoglm-zhipu-ai--xiao9905-github-io","name":"AutoGLM (Zhipu AI)","url":"https://xiao9905.github.io/AutoGLM/","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"GUI agent + cross-session task memory","license":"Apache-2.0","latest_release":"no releases","deployment":"Both","pricing":"Free + paid","mcp_support":"not documented publicly","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"zai-org/Open-AutoGLM — ~26.2k stars","storage":"file|vector","retrieval":"injection|similarity","persistence":"cross-session","update":"extraction","unit":"episode|fact","governance":"opaque","conflict":"n/a","description":"Tsinghua spinout. Foundation agent series for autonomous smartphone/PC GUI control. Persistent task memory retains shopping history, app usage patterns, multi-step task state across sessions. AutoGLM-Thinking claimed first agent capable of concurrent thinking and acting.","cited_cells":42,"third_party_cells":17,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/autoglm-zhipu-ai--xiao9905-github-io"},{"id":"autorag--arxiv-2410-20878","name":"AutoRAG","url":"https://arxiv.org/abs/2410.20878","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-29","type":"AutoML-style greedy RAG pipeline","license":"Apache-2.0","latest_release":"v0.3.27 (2026-10-04)","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"Marker-Inc-Korea/AutoRAG — ~5.1k stars, last commit n/a (PyPI release 2026-10)","storage":"hybrid","retrieval":"similarity|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Applies AutoML to RAG: greedy algorithm evaluates and selects best-performing module at each pipeline stage (query expansion, retrieval, reranking, generation). Validated on ARAGOG (423 papers, 107 QAs). The AutoML library now lives in the repo's legacy/ directory in maintenance mode (bug fixes, dependency updates, PyPI releases); the repo root now hosts AutoRAG 2.0, a personal-librarian agent with a desktop app.","cited_cells":28,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/autorag--arxiv-2410-20878"},{"id":"aws-bedrock-agentcore-memory--docs-aws-amazon-com","name":"AWS Bedrock AgentCore Memory","url":"https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/memory-integrate-lang.html","section":"Platform-provider memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Managed memory in Bedrock AgentCore","license":"","latest_release":"","deployment":"Managed-only","pricing":"Pay-per-use: short-term $1.00/GB ingested (was $0.25/1k events until 2026-10-05) + $0.10/GB-mo storage; long-term $0.75/1k records-mo (built-in) and $0.50/1k retrievals","mcp_support":"native (first-party) — AgentCore supports MCP","a2a_support":"searched not found","backend_storage":"custom (AWS-managed)","llm_lock":"","github":"","storage":"proprietary","retrieval":"similarity","persistence":"cross-session","update":"extraction","unit":"fact","governance":"opaque","conflict":"none","description":"AWS managed memory service for agents running on Bedrock AgentCore. Native integrations with LangChain/LangGraph and Strands. Drives long-running MCP servers + agent workloads.","cited_cells":39,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/aws-bedrock-agentcore-memory--docs-aws-amazon-com"},{"id":"backboard--github-com","name":"Backboard","url":"https://backboard.io/products/api","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-29","type":"Memory + LoCoMo evaluation harness","license":"Unlicensed","latest_release":"no releases","deployment":"Managed cloud API; vendor also advertises private-cloud, on-prem and on-device deployment","pricing":"Free $0; Education $5; Developer $20; Scale $400 (billing period not stated on page)","mcp_support":"not documented publicly","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"Backboard-io/Backboard-Locomo-Benchmark — ~18 stars (org's largest repo is Backboard-R-CLI, ~54 stars)","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"extraction","unit":"episode","governance":"inspectable","conflict":"none","description":"Stateful AI API combining persistent memory, RAG, web search and routing across many LLMs behind one interface, with SDKs for Python and npm; benchmarks its memory on LoCoMo and LongMemEval. As of Oct 2026 Backboard positions itself as an AI compute efficiency lab selling an inference-optimization engine; the memory product continues as its Stateful API.","cited_cells":38,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/backboard--github-com"},{"id":"basic-memory--gh-basicmachines-co-basic-memory","name":"Basic Memory","url":"https://github.com/basicmachines-co/basic-memory","section":"File-backed / editor paradigms","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"AGPL-3.0","latest_release":"last commit 2026-10-22","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/basicmachines-co/basic-memory (4322★)","storage":"file|relational|vector","retrieval":"hybrid-rerank|graph-traversal|agentic","persistence":"long-term","update":"agent-controlled|user-edit","unit":"document|fact|entity","governance":"editable|user-controllable","conflict":"manual","description":"Local-first MCP memory where each memory is a Markdown note with observations and wikilink relations that form a traversable knowledge graph, indexed with SQLite (optional Postgres vectors), plus an optional hosted cloud. widely used MCP memory server with a paid cloud tier","cited_cells":17,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/basic-memory--gh-basicmachines-co-basic-memory"},{"id":"beads-steve-yegge--gh-steveyegge-beads","name":"Beads (Steve Yegge)","url":"https://github.com/gastownhall/beads","section":"Coding-agent memory","tier":3,"status":"active","last_verified_at":"2026-10-29","type":"Framework","license":"MIT","latest_release":"last commit 2026-10-22","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/gastownhall/beads (moved from steveyegge/beads; ~28.3k★)","storage":"relational","retrieval":"exact-match|injection|graph-traversal","persistence":"long-term|project-bound","update":"agent-controlled|consolidation","unit":"document|fact|summary","governance":"auditable|editable","conflict":"branch-and-merge","description":"Dolt-backed, dependency-aware graph issue tracker that acts as persistent structured memory for coding agents, replacing markdown plans; `bd remember` stores durable facts and `bd prime` re-injects them.","cited_cells":16,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/beads-steve-yegge--gh-steveyegge-beads"},{"id":"bee--bee-computer","name":"Bee","url":"https://bee.computer","section":"Voice-first / wearable AI memory","tier":1,"status":"acquired","last_verified_at":"2026-10-29","type":"Audio pendant — never stores raw audio","license":"","latest_release":"","deployment":"Managed-only","pricing":"Bee Frontier bracelet $74.99 (sold via Amazon, launched 2026-10-15); Bee Pro subscription $11.99/mo (free tier keeps 14 days of memory); Bee Eclipse ($100) coming soon","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"lifelong","update":"append-only","unit":"episode","governance":"user-controllable","conflict":"append","description":"Acquired by Amazon (Jul 2025); bee.computer is now © Amazon and sells the Bee Frontier wearable ($74.99), with a Bee Eclipse device marked coming soon. Small wearable (pendant / clip / bracelet). Continuous capture; converts speech to text immediately; never stores raw audio . Daily summaries, personal fact sheet, to-do suggestions. Cloud processing on derived text only.","cited_cells":37,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/bee--bee-computer"},{"id":"benevolentai--benevolent-com","name":"BenevolentAI","url":"https://www.benevolent.com/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Continuously-refreshed biological KG","license":"","latest_release":"","deployment":"Managed-only","pricing":"Qor: free trial (Try Qor for free) plus demo-led enterprise sales; no public price list","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"extraction","unit":"fact","governance":"inspectable","conflict":"overwrite","description":"Taken private: merged into Osaka Holdings (renamed BenevolentAI S.A.) and delisted from Euronext Amsterdam 2025-03-13. Target identification / drug repurposing / mechanism tracing. 85+ data sources, petabyte-scale, rebuilt every few weeks. Wet-lab results re-enter the graph and shift downstream predictions — institutional experimental memory closing a feedback loop. As of Oct 2026 benevolent.com leads with Qor, powered by BenevolentAI, a life-science intelligence service for biopharma decision makers and consultancies. Qor (qor.one) is a continuously updated decision-intelligence service built on BenevolentAI's knowledge graph and reasoning methods: it tracks evidence on an asset or indication (competitive position, rival readouts, comparable deals), source-verifies each claim and flags when an earlier conclusion stops holding. It targets biotech teams and life-science consultancies.","cited_cells":33,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/benevolentai--benevolent-com"},{"id":"beyond-rag-for-agent-memory--arxiv-2602-02007","name":"Beyond RAG for Agent Memory","url":"https://arxiv.org/abs/2602.02007","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"Decoupled-aggregate memory retrieval","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"","storage":"hybrid","retrieval":"similarity|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"llm-arbitrate","description":"Argues standard RAG breaks on bounded agent conversation streams (highly correlated, redundant spans). Decomposes the stream into semantic units, organises them into themes, then inverts that structure for retrieval.","cited_cells":23,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/beyond-rag-for-agent-memory--arxiv-2602-02007"},{"id":"bge-m3--arxiv-2402-03216","name":"BGE-M3","url":"https://arxiv.org/abs/2402.03216","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"Unified dense-sparse-multi-vector embedding","license":"MIT","latest_release":"v1.4.2 (2026-08-24) — FlagEmbedding package","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"FlagOpen/FlagEmbedding — ~12.2k stars, last commit n/a (PyPI release 2026-08)","storage":"vector","retrieval":"similarity|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Single embedding model supporting dense, sparse, and multi-vector (ColBERT-style) retrieval simultaneously. 100+ languages, inputs up to 8192 tokens. Self-knowledge distillation across the three modes.","cited_cells":27,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/bge-m3--arxiv-2402-03216"},{"id":"bolt-new-stackblitz--bolt-new","name":"Bolt.new (StackBlitz)","url":"https://bolt.new","section":"Coding-agent memory","tier":2,"status":"stale","last_verified_at":"2026-10-28","type":"WebContainer file system","license":"MIT","latest_release":"no releases","deployment":"Managed-only","pricing":"Free 1M tokens/mo (300k/day); Pro $25/mo 10M tokens; Teams $30/member/mo; Enterprise custom","mcp_support":"no first-party MCP; WebContainer runtime restricts external server transports","a2a_support":"searched not found","backend_storage":"custom (browser-based WebContainer)","llm_lock":"","github":"stackblitz/bolt.new — ~16.6k stars, last commit 2024-12 (open-source repo, not archived)","storage":"file","retrieval":"injection","persistence":"session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"n/a","description":"In-browser WebContainer (Rust-backed shared memory) is the implicit state store. No dedicated memory abstraction — project files are the state. Team plan adds shared workspaces + GitHub integration.","cited_cells":46,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/bolt-new-stackblitz--bolt-new"},{"id":"botpress-llmz--botpress-com","name":"Botpress LLMz","url":"https://botpress.com/features/knowledge-bases","section":"Framework-embedded memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Vector DB + LLMz engine + KB","license":"MIT (SDK/integration packages in botpress/botpress)","latest_release":"","deployment":"Both (OSS self-hosted; Botpress Cloud SaaS; private cloud option)","pricing":"Free 500 msgs/mo; Plus $89/mo 1GB Vector DB; Team $495/mo 2GB; Enterprise custom","mcp_support":"via official adapter — Botpress MCP integration","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"multi-vendor","github":"botpress/botpress — ~14.9k stars","storage":"vector","retrieval":"similarity|injection","persistence":"cross-session","update":"extraction","unit":"document|episode","governance":"inspectable","conflict":"llm-arbitrate","description":"Per-plan vector-DB storage quota + LLMz autonomous engine (in-session working memory) + Knowledge Base (semantic search over uploaded docs). Long-term user memory persists across sessions.","cited_cells":47,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/botpress-llmz--botpress-com"},{"id":"brave-leo--brave-com","name":"Brave Leo","url":"https://brave.com/leo/","section":"Browser-agent memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"User-curated local-only memory","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free + paid","mcp_support":"searched not found — Leo focuses on browser-local context, no MCP surfaced","a2a_support":"searched not found","backend_storage":"","llm_lock":"","github":"","storage":"file","retrieval":"injection","persistence":"long-term","update":"user-controllable","unit":"fact","governance":"user-controllable","conflict":"none","description":"Privacy-focused integrated assistant. Memory is user-curated: explicit facts and preferences persist across all sessions, stored locally with no server retention. Agentic browsing experiment (Brave Nightly Dec 2025) takes multi-step actions inside sites referencing those preferences.","cited_cells":38,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/brave-leo--brave-com"},{"id":"browserbase--browserbase-com","name":"Browserbase","url":"https://www.browserbase.com/","section":"Browser-agent memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Programmatic session persistence (dev infra)","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free $0; Developer $20/mo; Startup $99/mo; Scale custom","mcp_support":"native (first-party) — Browserbase MCP","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"file","retrieval":"extraction-pull","persistence":"session","update":"overwrite","unit":"episode","governance":"inspectable","conflict":"none","description":"Cloud browser infrastructure for AI agents. Headless Chromium sessions with persistent cookies, localStorage, session recording — agent workflows carry state across runs. Memory at the session/automation tier rather than conversational.","cited_cells":64,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/browserbase--browserbase-com"},{"id":"byterover--byterover-dev","name":"ByteRover","url":"https://www.byterover.dev/","section":"Dedicated memory layers","tier":2,"status":"archived","last_verified_at":"2026-10-29","type":"Agent memory product","license":"Elastic-2.0 (CLI repo, now archived)","latest_release":"","deployment":"Local-first CLI (curl installer or npm byterover-cli; background daemon since v2.0.0); optional cloud for team sync, multi-machine access and backup","pricing":"Free $0; Pro $14.90/mo; Team $149/mo (third-party listing checked 2026-08-26)","mcp_support":"native (first-party) — ByteRover MCP","a2a_support":"not documented publicly","backend_storage":"custom","llm_lock":"","github":"campfirein/byterover-cli — ~5k stars, archived 2026-09-27","storage":"hybrid","retrieval":"similarity","persistence":"long-term","update":"extraction","unit":"episode","governance":"user-controllable","conflict":"none","description":"Agent-memory product. LLM-curated hierarchical context. Now ships as ByteRover Desktop (V4) with cloud-synced spaces for agents and teammates; the open-source CLI repo campfirein/byterover-cli is archived.","cited_cells":33,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/byterover--byterover-dev"},{"id":"cag-cache-augmented-generation--arxiv-2412-15605","name":"CAG (Cache-Augmented Generation)","url":"https://arxiv.org/abs/2412.15605","section":"Retrieval-as-memory hybrids","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"KV-cache preloading replaces retrieval","license":"MIT","latest_release":"no-release no-release","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"hhhuang/CAG — ~1.6k stars, last commit n/a","storage":"kv-cache","retrieval":"cache-lookup","persistence":"session","update":"read-only","unit":"document","governance":"inspectable","conflict":"n/a","description":"\"Don't Do RAG.\" Preloads all relevant documents into context, caches the resulting KV state; at inference the model answers from cache with no retrieval step — eliminates retrieval latency + retrieval-error failure modes.","cited_cells":29,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cag-cache-augmented-generation--arxiv-2412-15605"},{"id":"capacities--capacities-io","name":"Capacities","url":"https://capacities.io/","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Object-graph PKM (typed objects)","license":"","latest_release":"","deployment":"Managed-only","pricing":"Basic free; Pro $9.99/mo (16% off yearly); Believer from $12.49/mo; AI Plus add-on $8.33/mo (annual)","mcp_support":"unclear — Pro/Believer list 'AI Chat Connectors' for ChatGPT, Claude and Cursor; MCP not named on pricing page","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph|relational","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"overwrite","unit":"document|fact","governance":"user-controllable","conflict":"overwrite","description":"Object-graph (people, books, projects as typed objects); daily notes; bidirectional links. AI chat grounded in current note; AI property auto-fill; in-editor writing assistant; pattern-finding across objects.","cited_cells":36,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/capacities--capacities-io"},{"id":"causaly--causaly-com","name":"Causaly","url":"https://www.causaly.com/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Persistent causal knowledge graph","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"yes — Causaly capabilities exposed via MCP, A2A and APIs","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph","retrieval":"graph-traversal","persistence":"long-term","update":"extraction|overwrite","unit":"fact","governance":"inspectable","conflict":"overwrite","description":"Agentic AI platform for life sciences: orchestrated scientific agents over a curated biomedical knowledge graph (plus Bio Graph and Pipeline Graph), an evidence-retrieval system and an enterprise data fabric over literature, trials and patents.","cited_cells":35,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/causaly--causaly-com"},{"id":"cellon-memory-inc--en-wowtale-net","name":"Cellon (Memory.inc)","url":"https://en.wowtale.net/2026/04/30/233992/","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-29","type":"Cross-service context stitching / retrieval memory","license":"","latest_release":"","deployment":"searched not found","pricing":"searched not found","mcp_support":"not documented publicly","a2a_support":"not documented publicly","backend_storage":"custom","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"cross-session","update":"extraction","unit":"fact|episode","governance":"opaque","conflict":"none","description":"Korean startup. Connects scattered internal/external data across multiple chats and services to improve AI answer quality at retrieval. 18-year-old CEO Jang Jun-ha; pitch is that answer quality depends on what the model remembers, not which model. As of Oct 2026 memory.inc redirects to Pogl (about.pogl.com), an AI game and experience maker from the same company (Celon Inc., CEO Junha Jang); the memory product is no longer presented.","cited_cells":26,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cellon-memory-inc--en-wowtale-net"},{"id":"character-ai--character-ai","name":"Character.ai","url":"https://character.ai","section":"Vertical / domain-specific AI memory","tier":1,"status":"dead","last_verified_at":"2026-10-29","type":"Layered: chat memories + auto-memories + pinned","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free + paid","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"kv|vector","retrieval":"injection|similarity","persistence":"long-term","update":"extraction|overwrite","unit":"fact|episode","governance":"user-controllable","conflict":"overwrite","description":"Chat Memories (user-defined facts), auto-memories for c.ai+ subscribers, pinned memories, in-context retention. Latest chat models PipSqueak 3 and ShortSqueak (Sep 2026). On 2026-10-12 the company announced a voluntary Chapter 11 process for a court-supervised sale of all or substantially all assets (sale expected late Nov/early Dec 2026); the platform keeps running meanwhile and would be wound down if no viable bid emerges.","cited_cells":60,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/character-ai--character-ai"},{"id":"charisma-ai--charisma-ai","name":"Charisma.ai","url":"https://charisma.ai","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Author-controlled memory nodes + lying / misremembering","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free + paid","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph","retrieval":"graph-traversal|injection","persistence":"long-term","update":"overwrite","unit":"fact|episode","governance":"user-controllable","conflict":"overwrite","description":"No-code interactive narrative platform. Characters track what they know, when they were lied to, what they misremember; story state changes downstream behaviour. Characters can deliberately lie / misremember and track when players have lied.","cited_cells":49,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/charisma-ai--charisma-ai"},{"id":"chatgpt-study-mode--openai-com","name":"ChatGPT Study Mode","url":"https://openai.com/index/chatgpt-study-mode/","section":"Vertical / domain-specific AI memory","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Socratic mode + general ChatGPT Memory","license":"","latest_release":"","deployment":"Web + iOS + Android","pricing":"Available on Free, Go ($8/mo), Plus ($20/mo), Pro ($200/mo), Team/Business","mcp_support":"yes (platform-level) — ChatGPT custom MCP connectors via Developer Mode on web for Business/Enterprise/Edu; Pro read-only","a2a_support":"","backend_storage":"","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"cross-session","update":"extraction","unit":"fact","governance":"user-controllable","conflict":"llm-arbitrate","description":"Activates a Socratic interaction style — guiding questions, scaffolded concepts, knowledge checks rather than direct answers. With ChatGPT Memory enabled, persists learning goals and facts across sessions. Memory is not study-mode-specific; it is the same general ChatGPT Memory, so personalization depth depends on whether the user has memory enabled and populated.","cited_cells":32,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/chatgpt-study-mode--openai-com"},{"id":"chunkrag--arxiv-2410-19572","name":"ChunkRAG","url":"https://arxiv.org/abs/2410.19572","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"LLM-scored chunk-level filtering","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"","storage":"vector","retrieval":"similarity|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Semantic chunking divides documents into coherent sections; LLM-based relevance scorer filters each chunk against the query before generation, pruning irrelevant chunks and reducing hallucination.","cited_cells":22,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/chunkrag--arxiv-2410-19572"},{"id":"cipher-byterover--gh-campfirein-cipher","name":"Cipher (byterover) (now ByteRover CLI)","url":"https://github.com/campfirein/byterover-cli","section":"Dedicated memory layers","tier":2,"status":"archived","last_verified_at":"2026-10-29","type":"Cross-IDE memory layer (vector + KG)","license":"Elastic-2.0","latest_release":"v3.16.1 (2026-05-27)","deployment":"Local CLI (works offline with own/local LLM provider); optional ByteRover cloud for team sync and backup","pricing":"searched not found","mcp_support":"native (first-party MCP server)","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"campfirein/byterover-cli (formerly cipher) — ~5.0k stars, archived 2026-09","storage":"vector|graph","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction","unit":"fact","governance":"opaque","conflict":"n/a","description":"Renamed ByteRover CLI (brv); repo campfirein/byterover-cli archived read-only 2026-09-27, license changed to Elastic License 2.0. Opensource memory layer focused on programming agents — works with Cursor, Windsurf, Claude Code, Roo Code via MCP. Maintains a coding-context memory across sessions and across IDEs.","cited_cells":17,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cipher-byterover--gh-campfirein-cipher"},{"id":"claude-brain--gh-memvid-claude-brain","name":"claude-brain","url":"https://github.com/memvid/claude-brain","section":"Claude Code memory mechanisms","tier":2,"status":"active","last_verified_at":"2026-10-23","type":"Memvid → Claude Code plugin","license":"MIT","latest_release":"v1.0.7 (2026-01-05)","deployment":"local","pricing":"free / open source","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"memvid/claude-brain — ~575 stars, last commit not confirmed","storage":"file","retrieval":"similarity","persistence":"long-term","update":"append-only","unit":"chunk","governance":"inspectable","conflict":"append","description":"Claude Code plugin packaging of Memvid (see Dedicated memory layers). Wraps the Memvid .mv2 single-file store as a slash-command-installable plugin so per-project memory becomes a git-committable file inside the repo.","cited_cells":43,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/claude-brain--gh-memvid-claude-brain"},{"id":"claude-code-auto-memory--docs-claude-com","name":"Claude Code auto-memory","url":"https://docs.claude.com/en/docs/claude-code/memory","section":"Claude Code memory mechanisms","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Per-project file directory","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"native (first-party) — Claude Code consumes MCP","a2a_support":"not supported","backend_storage":"file (~/.claude/projects/<project>/MEMORY.md)","llm_lock":"","github":"","storage":"file","retrieval":"injection|extraction-pull","persistence":"cross-session","update":"append-only|overwrite","unit":"fact|document","governance":"user-controllable","conflict":"append","description":"Persistent file-based memory at ~/.claude/projects/<project>/memory/ with a MEMORY.md index plus typed memory files (user / feedback / project / reference). Loaded into context across sessions.","cited_cells":55,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/claude-code-auto-memory--docs-claude-com"},{"id":"claude-cursormemorymcp--gh-angleito-claude-cursormemorymcp","name":"Claude-CursorMemoryMCP","url":"https://github.com/Angleito/Claude-CursorMemoryMCP","section":"Claude Code memory mechanisms","tier":2,"status":"stale","last_verified_at":"2026-10-28","type":"PostgreSQL + pgvector with RBAC","license":"Unlicensed","latest_release":"no releases","deployment":"local","pricing":"free / open source","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"Angleito/Claude-CursorMemoryMCP — 4 stars, last commit 2025-05","storage":"vector|relational","retrieval":"similarity","persistence":"long-term","update":"overwrite|append-only","unit":"chunk","governance":"auditable","conflict":"llm-arbitrate","description":"FastAPI server. RBAC + API-key auth, encryption, backup/recovery, embedding pipeline with deduplication. Only server in this list with role-based access control over pgvector.","cited_cells":41,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/claude-cursormemorymcp--gh-angleito-claude-cursormemorymcp"},{"id":"claude-md--docs-claude-com","name":"CLAUDE.md","url":"https://docs.claude.com/en/docs/claude-code/memory","section":"File-backed / editor paradigms","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Git-backed persistent instructions","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"file (CLAUDE.md)","llm_lock":"","github":"","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"manual","description":"CLAUDE.md files (managed, user, project, local scopes) concatenated into context at session start, with @imports and path-scoped .claude/rules/; complemented by Claude-written auto memory (MEMORY.md index). Project-root file re-injected after /compact.","cited_cells":50,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/claude-md--docs-claude-com"},{"id":"claude-mem--gh-thedotmack-claude-mem","name":"claude-mem","url":"https://github.com/thedotmack/claude-mem","section":"Claude Code memory mechanisms","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Episodic / session-log compression","license":"Apache-2.0","latest_release":"v15.0.1 (2026-10-20)","deployment":"local — SQLite + FTS5 per project; optional Chroma vector; no cloud","pricing":"Open-source engine free (Apache-2.0, local memory); CMEM Pro hosted observation + cloud sync + private MCP recall: Pro $30/mo, Heavy $70/mo, CMEM Max $100/mo (with billed overage), 30-day free trial; TeamBrain team pricing not announced (pilot)","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"thedotmack/claude-mem — ~104k stars, last commit 2026-10","storage":"file","retrieval":"injection","persistence":"cross-session","update":"consolidation","unit":"summary|episode","governance":"inspectable","conflict":"append","description":"Hooks into Claude Code's session lifecycle; compresses tool outputs into ~500-token observations via Claude's agent SDK. SQLite + FTS5 with optional Chroma vector. \"Endless Mode\" compresses in-flight to extend session length. Captures what the agent did , not just what the user typed.","cited_cells":66,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/claude-mem--gh-thedotmack-claude-mem"},{"id":"claude-supermemory--gh-supermemoryai-claude-supermemory","name":"claude-supermemory","url":"https://github.com/supermemoryai/claude-supermemory","section":"Claude Code memory mechanisms","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Team-scope memory plugin","license":"MIT","latest_release":"v0.0.2 2026-02-09","deployment":"Supermemory cloud API by default (API key); local self-host option via npx supermemory local","pricing":"requires Supermemory Pro (pricing not publicly listed; contact sales)","mcp_support":"no — implemented as Claude Code plugin hooks (Node.js scripts), not an MCP server","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"supermemoryai/claude-supermemory — ~2.8k stars","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"extraction","unit":"fact|document","governance":"opaque","conflict":"llm-arbitrate","description":"Wraps Supermemory cloud API. Team memory shared across project members, separate from personal memories. Auto-capture on session end; per-repo config.","cited_cells":38,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/claude-supermemory--gh-supermemoryai-claude-supermemory"},{"id":"cline-clinerules--cline-bot","name":"Cline .clinerules","url":"https://cline.bot/blog/clinerules-version-controlled-shareable-and-ai-editable-instructions","section":"File-backed / editor paradigms","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Version-controlled, AI-editable rules","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"file","llm_lock":"","github":"","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"manual","description":"Markdown rule files in a workspace .clinerules/ (or .cline/rules/) directory plus global rules in ~/Documents/Cline/Rules; workspace wins on conflict. Per-rule toggles, conditional rules via YAML 'paths' globs, and AGENTS.md / .cursorrules / .windsurfrules are also read.","cited_cells":30,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cline-clinerules--cline-bot"},{"id":"cline-memory-bank--docs-cline-bot","name":"Cline Memory Bank","url":"https://docs.cline.bot/features/memory-bank","section":"File-backed / editor paradigms","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Session-spanning state pattern","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"file (memory-bank/ markdown)","llm_lock":"","github":"","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"manual","description":"memory-bank/ directory of Markdown: projectBrief.md , productContext.md , activeContext.md , systemPatterns.md , techContext.md , progress.md . Loaded in full at start of every task. Compensates for context-window reset between sessions.","cited_cells":29,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cline-memory-bank--docs-cline-bot"},{"id":"clio-duo-manage-ai--clio-com","name":"Clio Duo / Manage AI","url":"https://www.clio.com/features/legal-ai-software/","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Matter-scoped PMS-native context retrieval","license":"","latest_release":"","deployment":"Cloud SaaS only (runs entirely in Clio's cloud infrastructure)","pricing":"from $49/user/mo (third-party listing; AI add-on pricing not public)","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|relational","retrieval":"similarity|injection","persistence":"cross-session","update":"extraction","unit":"document|fact","governance":"auditable","conflict":"document-version","description":"SMB law firms. AI embedded in practice management; pulls full matter overviews — WIP, billing, documents, communications — as context. Memory is the PMS data; no separately persisted cross-matter layer. Renamed Manage AI at ClioCon 2025.","cited_cells":28,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/clio-duo-manage-ai--clio-com"},{"id":"cloudflare-agent-memory--infoq-com","name":"Cloudflare Agent Memory","url":"https://blog.cloudflare.com/introducing-agent-memory/","section":"Platform-provider memory","tier":2,"status":"active","last_verified_at":"2026-10-29","type":"Managed persistent memory service","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free during private beta (not billed; Cloudflare promises 30 days' notice before charging)","mcp_support":"native (first-party) — Cloudflare Agents SDK serves MCP","a2a_support":"searched not found","backend_storage":"custom (Durable Objects + Vectorize)","llm_lock":"","github":"","storage":"hybrid","retrieval":"similarity","persistence":"cross-session","update":"agent-controlled","unit":"episode","governance":"opaque","conflict":"none","description":"Managed persistent memory service for AI agents on Cloudflare's edge. Private beta announced 2026-04-17. Sits alongside Workers AI / Vectorize as the memory tier of the Cloudflare agent stack.","cited_cells":39,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cloudflare-agent-memory--infoq-com"},{"id":"cocounsel-legal-thomson-reuters--legaltechnology-com","name":"CoCounsel Legal (Thomson Reuters)","url":"https://legal.thomsonreuters.com/en/products/cocounsel-legal","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-29","type":"Session-scoped agentic context","license":"","latest_release":"","deployment":"Managed SaaS; GA in US, Canada/UK/Australia slated for later 2026","pricing":"searched not found","mcp_support":"yes — first-party CoCounsel Legal MCP connector for Claude (May 2026, expanded Aug 2026); AWS MCP connection announced","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"session","update":"overwrite","unit":"document","governance":"auditable","conflict":"none","description":"Next-gen release (Aug 2026, built on Claude Agent SDK) is fully agentic and adds per-matter Workspaces that keep firm documents and precedents as context across matters and teams, plus Brief Builder, a Word drafting agent and Deep Research Verify.","cited_cells":31,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cocounsel-legal-thomson-reuters--legaltechnology-com"},{"id":"codebase-memory-mcp-deusdata--gh-deusdata-codebase-memory-mcp","name":"codebase-memory-mcp (DeusData)","url":"https://github.com/DeusData/codebase-memory-mcp","section":"Claude Code memory mechanisms","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Tree-sitter knowledge-graph MCP for code","license":"MIT","latest_release":"v0.11.0 (2026-09-15)","deployment":"Local — single static binary (macOS arm64/amd64; Linux arm64/amd64; Windows amd64)","pricing":"free / open source (MIT); no paid tier","mcp_support":"native (first-party MCP server)","a2a_support":"not supported","backend_storage":"SQLite (WAL mode) at ~/.cache/codebase-memory-mcp/","llm_lock":"","github":"DeusData/codebase-memory-mcp — ~43.3k stars, last release 2026-09","storage":"graph","retrieval":"structural-query","persistence":"project-bound","update":"replacement","unit":"fact","governance":"user-controllable","conflict":"deterministic-overwrite","description":"High-performance MCP server building persistent code knowledge graphs via AST. 155 languages via vendored tree-sitter; SQLite + WAL; single static binary; zero runtime deps. Knowledge graph: Project/Package/File/Module/Class/Function/Method/Interface/Enum/Type/Route/Resource nodes; CALLS/IMPORTS/DEFINES/IMPLEMENTS/INHERITS/HTTP_CALLS/ASYNC_CALLS/EMITS/LISTENS_ON/DATA_FLOWS/SIMILAR_TO/SEMANTICALLY_RELATED edges. 14 MCP tools incl. Cypher-like query.","cited_cells":56,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/codebase-memory-mcp-deusdata--gh-deusdata-codebase-memory-mcp"},{"id":"cognee--cognee-ai","name":"Cognee","url":"https://www.cognee.ai/","section":"Dedicated memory layers","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Knowledge graph + ECL pipeline","license":"Apache-2.0","latest_release":"v1.6.5 (2026-10-16)","deployment":"Both (OSS self-hosted; Cognee Cloud managed; Enterprise BYOC)","pricing":"Free $0 (1 workspace, 1M tokens); Standard $1 per 1M tokens + $5/workspace/mo; Enterprise BYOC custom","mcp_support":"native (first-party) — cognee-mcp","a2a_support":"not documented publicly","backend_storage":"hybrid (vector + graph)","llm_lock":"","github":"topoteretes/cognee — ~31.2k stars, last commit 2026-10","storage":"graph","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"extraction","unit":"fact|document","governance":"inspectable","conflict":"llm-arbitrate","description":"\"Extract–Cognify–Load\" pipeline that turns raw input into a typed, queryable knowledge graph for agent recall.","cited_cells":71,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cognee--cognee-ai"},{"id":"cognee-mcp--gh-topoteretes-cognee","name":"Cognee MCP","url":"https://github.com/topoteretes/cognee","section":"Claude Code memory mechanisms","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Multi-layer KG with rule extraction","license":"Apache-2.0","latest_release":"v1.6.5 (2026-10-16)","deployment":"self-hosted (local vector + graph DB); cloud option exists (topoteretes managed)","pricing":"OSS (Apache-2.0) self-hosted; Cognee Cloud managed service (pricing not on README); some production components (Postgres graph store) are licensed products","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"hybrid (vector + graph)","llm_lock":"","github":"topoteretes/cognee — ~31.2k stars, last commit 2026-10","storage":"graph","retrieval":"graph-traversal","persistence":"long-term","update":"extraction","unit":"fact","governance":"inspectable","conflict":"llm-arbitrate","description":"\"Cognify\" extracts entities and relationships into a structured graph. Saves interaction patterns as coding rules; uses both vector and graph DBs. Bundled in Cognee v0.3.5+.","cited_cells":45,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cognee-mcp--gh-topoteretes-cognee"},{"id":"cognigy--cognigy-com","name":"Cognigy (now NiCE Cognigy)","url":"https://www.cognigy.com/product-updates/an-ultimate-guide-to-ai-agent-memory","section":"Vertical / domain-specific AI memory","tier":1,"status":"acquired","last_verified_at":"2026-10-23","type":"Two-tier short + long-term, shared across agents","license":"","latest_release":"","deployment":"Both","pricing":"Enterprise only","mcp_support":"yes — MCP client via MCP Node in Cognigy.AI (since 2025-04); standalone @cognigy/mcp-server npm package now deprecated in favour of the NiCE Cognigy plugin","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"kv|vector","retrieval":"exact-match|similarity|injection","persistence":"cross-session","update":"append-only|overwrite","unit":"episode|fact","governance":"inspectable","conflict":"none","description":"Acquired by NiCE (closed 2025-09-08); continues as NiCE Cognigy, standalone and inside CXone Mpower. Enterprise contact centre (voice + chat). Short-term covers current session; long-term persists across sessions and is shared by every agent a customer might encounter. Most explicitly engineered memory model in the segment.","cited_cells":36,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cognigy--cognigy-com"},{"id":"context-mode--gh-mksglu-context-mode","name":"context-mode","url":"https://github.com/mksglu/context-mode","section":"Coding-agent memory","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"Elastic License 2.0","latest_release":"last commit 2026-10-22","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/mksglu/context-mode (31791★)","storage":"relational","retrieval":"exact-match|injection","persistence":"session|cross-session","update":"evict-oldest|consolidation","unit":"episode|chunk","governance":"user-controllable|deterministic","conflict":"append-only","description":"MCP server + hooks that sandbox tool output out of the coding agent's context window (claims ~98% reduction) and persist session memory across 17 agent platforms.","cited_cells":16,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/context-mode--gh-mksglu-context-mode"},{"id":"continue-dev-continue-rules--docs-continue-dev","name":"Continue.dev .continue/rules","url":"https://docs.continue.dev/customize/deep-dives/rules","section":"File-backed / editor paradigms","tier":2,"status":"archived","last_verified_at":"2026-10-29","type":"Path-glob-scoped rules","license":"Apache-2.0","latest_release":"v2.0.0-vscode (2026-06-19) — final release","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"file","llm_lock":"","github":"continuedev/continue — ~35.7k stars, last commit 2026-06 (read-only, unmaintained)","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"manual","description":".continue/rules/*.md at workspace root, loaded lexicographically; YAML frontmatter name, globs, regex, description, alwaysApply. Continue was acquired by Cursor in June 2026 and the extension is no longer maintained.","cited_cells":34,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/continue-dev-continue-rules--docs-continue-dev"},{"id":"continue-dev-memory-mcp--hub-continue-dev","name":"Continue.dev Memory MCP","url":"https://hub.continue.dev/continuedev/memory-mcp","section":"Coding-agent memory","tier":2,"status":"archived","last_verified_at":"2026-10-29","type":"KG-backed memory via MCP","license":"","latest_release":"","deployment":"Both (VS Code/JetBrains extension local + optional shared MCP server)","pricing":"Discontinued — Continue acquired by Cursor; recurring billing disabled and unused paid credits refunded; Continue Hub (hub.continue.dev) offline","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph","retrieval":"graph-traversal","persistence":"long-term","update":"extraction","unit":"fact|document","governance":"inspectable","conflict":"none","description":"Discontinued: Continue was bought by Cursor (announced ~2026-06-16); the product is discontinued and the continuedev/continue repo is read-only. Official Memory MCP (Mar 2025) backed by a knowledge graph. Memory is injected via MCP rather than native to the Continue core. Scope depends on MCP server deployment (local or shared).","cited_cells":36,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/continue-dev-memory-mcp--hub-continue-dev"},{"id":"convai-mimir--convai-com","name":"Convai (Mimir)","url":"https://convai.com","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"3-tier hierarchical memory + scene awareness","license":"","latest_release":"","deployment":"Both","pricing":"Credit-based subscriptions (~1,000 credits per $); e.g. Professional $99/mo = 99k credits; full tier list not retrievable","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|kv","retrieval":"similarity|injection","persistence":"long-term","update":"consolidation|extraction","unit":"episode|summary|fact","governance":"opaque","conflict":"none","description":"Mimir hierarchical memory: short-term (verbatim), medium-term (summaries + emotion), long-term (cross-session). Scene awareness layer adds in-engine perception. Unity + Unreal SDKs with real-time voice + lipsync.","cited_cells":29,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/convai-mimir--convai-com"},{"id":"core-redplanet--gh-redplanethq-core","name":"CORE (RedPlanet)","url":"https://github.com/RedPlanetHQ/core","section":"Personal AI / PKM / lifelogging memory","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Product","license":"AGPL-3.0","latest_release":"last commit 2026-09-07","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/RedPlanetHQ/core (2003★)","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"extraction|append-only","unit":"fact|episode|entity","governance":"auditable|user-controllable","conflict":"bi-temporal","description":"Always-on 'personal AI OS' that indexes email, meetings, GitHub, Linear, Slack and assistant chats into a temporal knowledge graph with fact classification and intent-driven retrieval, plus MCP integrations. reports 88.24% on LoCoMo (vendor claim)","cited_cells":14,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/core-redplanet--gh-redplanethq-core"},{"id":"craft-ai--craft-do","name":"Craft AI","url":"https://www.craft.do/","section":"Personal AI / PKM / lifelogging memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Block-based docs + AI assistant","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free; Plus $8/mo ($6.40/mo annual); Family $15/mo (2–6 accounts); Team $50/mo (up to 10 Plus accounts)","mcp_support":"yes — 'API & MCP access' included on Free and Plus tiers","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"","storage":"relational","retrieval":"injection|similarity","persistence":"long-term","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"overwrite","description":"Hierarchical block-based documents + nested pages (no native graph). In-document AI assistant: summarise, rewrite, translate, Q&A over docs, proofread.","cited_cells":29,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/craft-ai--craft-do"},{"id":"crag--arxiv-2401-15884","name":"CRAG","url":"https://arxiv.org/abs/2401.15884","section":"Retrieval-as-memory hybrids","tier":4,"status":"research-code","last_verified_at":"2026-10-22","type":"Retrieval self-correction with web fallback","license":"no-license","latest_release":"no-release no-release","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"HuskyInSalt/CRAG — ~469 stars, last update noted 2024-10","storage":"vector","retrieval":"similarity|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Lightweight retrieval evaluator scores docs and triggers different actions: use, supplement with web search, or discard. Decompose-then-recompose algorithm filters irrelevant content from retrieved passages.","cited_cells":26,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/crag--arxiv-2401-15884"},{"id":"cresta--cresta-com","name":"Cresta","url":"https://cresta.com","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Shared memory across human/AI handoffs","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|kv","retrieval":"similarity|injection","persistence":"cross-session","update":"extraction","unit":"episode|fact","governance":"inspectable","conflict":"none","description":"Sales + contact centre. Customer history feeds real-time guidance and coaching. Memory survives handoffs between AI agent, human agent, and channels — but leans on CRM as the primary store.","cited_cells":29,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cresta--cresta-com"},{"id":"crewai-memory--crewai-com","name":"CrewAI Memory","url":"https://www.crewai.com/","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Short-term + long-term + entity","license":"MIT","latest_release":"v1.15.28 (2026-10-14)","deployment":"Both","pricing":"Free + paid","mcp_support":"via official adapter — crewai-tools MCP","a2a_support":"searched not found","backend_storage":"pluggable (ChromaDB default)","llm_lock":"any (BYO)","github":"50.7k★ +1.6k/mo Python","storage":"vector|kv","retrieval":"similarity|injection","persistence":"long-term","update":"extraction","unit":"episode|fact","governance":"inspectable","conflict":"llm-arbitrate","description":"Memory subsystem inside the CrewAI orchestration framework; integrates with Mem0 for the long-term tier.","cited_cells":54,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/crewai-memory--crewai-com"},{"id":"cursor-memories--docs-cursor-com","name":"Cursor Memories","url":"https://docs.cursor.com/context/memories","section":"Coding-agent memory","tier":3,"status":"active","last_verified_at":"2026-10-29","type":"Product","license":"proprietary","latest_release":"searched not found","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"","storage":"proprietary","retrieval":"injection","persistence":"cross-session|project-bound","update":"extraction|agent-controlled","unit":"fact","governance":"editable","conflict":"n/a","description":"Project-scoped memories in Cursor, created either by a background 'sidecar' model watching chats or by the agent via tool calls, and managed as auto-generated rules in settings. Traction: First-party feature of one of the most widely used AI IDEs. As of 2026-10-29 the Memories docs page redirects to the Cursor docs home, which no longer documents Memories; Rules remain the documented persistent-context mechanism.","cited_cells":13,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cursor-memories--docs-cursor-com"},{"id":"cursor-rules--docs-cursor-com","name":"Cursor Rules","url":"https://docs.cursor.com/context/rules","section":"File-backed / editor paradigms","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"IDE-embedded code memory","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"file (.cursor/rules/)","llm_lock":"","github":"","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"n/a","description":"Persistent instructions injected at start of model context: Project Rules (.mdc files in .cursor/rules, scoped Always/Auto Attached/Agent Requested/Manual), global User Rules, and AGENTS.md; legacy .cursorrules deprecated. Codebase indexing is separate from rules.","cited_cells":51,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/cursor-rules--docs-cursor-com"},{"id":"decagon--decagon-ai","name":"Decagon","url":"https://decagon.ai","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-23","type":"Cross-channel conversation graph","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph","retrieval":"graph-traversal","persistence":"long-term","update":"extraction","unit":"episode","governance":"inspectable","conflict":"none","description":"Autonomous support agents resolving end-to-end (refunds, cancellations). Channel-agnostic memory: chat → call carries prior context. Agents recognise returning customers and recall prior issues.","cited_cells":56,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/decagon--decagon-ai"},{"id":"deepscribe--deepscribe-ai","name":"DeepScribe","url":"https://www.deepscribe.ai/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Longitudinal oncology pull-forward","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|relational","retrieval":"similarity|injection","persistence":"long-term","update":"extraction","unit":"episode|document","governance":"auditable","conflict":"editor-in-the-loop","description":"Specialty-care ambient scribe (oncology focus). Pulls forward prior notes with precise dates; synthesises interval history (what changed since last visit); auto-populates disease-status and treatment-plan context.","cited_cells":36,"third_party_cells":8,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/deepscribe--deepscribe-ai"},{"id":"devin-cognition--devin-ai","name":"Devin (Cognition)","url":"https://devin.ai","section":"Coding-agent memory","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Server-side Knowledge + auto-Repo Knowledge","license":"","latest_release":"","deployment":"Both","pricing":"Free $0; Pro $20/mo; Max $200/mo; Teams $80/mo; Enterprise custom","mcp_support":"via official adapter — Devin supports MCP servers","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"kv|file","retrieval":"injection|extraction-pull","persistence":"cross-session","update":"agent-controlled","unit":"fact|document","governance":"inspectable","conflict":"none","description":"Structured Knowledge items stored server-side. Three-tier scope: repo-pinned (always injected), global (recall by relevance), org/enterprise (visible to org members). Auto-generates Repo Knowledge from repo structure each session; reads past session trajectories for self-improvement; writes notes for recurring scheduled tasks.","cited_cells":69,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/devin-cognition--devin-ai"},{"id":"dia-atlassian--diabrowser-com","name":"Dia (Atlassian)","url":"https://www.diabrowser.com","section":"Browser-agent memory","tier":1,"status":"acquired","last_verified_at":"2026-10-23","type":"Local-first browser-history + @history queries","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free (no AI); paid AI tiers 'Better Answers' $20/mo (page also shows $10) and 'Better Days' $100/mo","mcp_support":"Skills can wrap MCP servers via Atlassian Forge bridge; not native","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"file","retrieval":"extraction-pull|injection","persistence":"long-term","update":"append-only","unit":"episode|document","governance":"user-controllable","conflict":"none","description":"The Browser Company's AI-native browser; Atlassian acquired The Browser Company for $610M cash (announced 2025-09, closed 2025-10-21). Stores summaries and preferences locally; opt-in @history feature references up to 7 days of browsing as context. History/chats/bookmarks encrypted local-first.","cited_cells":59,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/dia-atlassian--diabrowser-com"},{"id":"dspy-history--dspy-ai","name":"DSPy History","url":"https://dspy.ai/api/primitives/History/","section":"Framework-embedded memory","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Signature-field / prompt-injection contract","license":"MIT","latest_release":"","deployment":"Both (OSS self-hosted; dspy.ai cloud services if applicable)","pricing":"OSS free (Apache-2.0); dspy.ai commercial services ~$110K ARR but pricing details not publicly listed","mcp_support":"searched not found","a2a_support":"not supported","backend_storage":"","llm_lock":"any (BYO via dspy.LM)","github":"stanfordnlp/dspy — ~36.7k stars, last commit n/a","storage":"kv","retrieval":"injection","persistence":"session","update":"read-only","unit":"document","governance":"inspectable","conflict":"n/a","description":"dspy.History primitive — typed field holding messages: list[dict] that slots into any Signature . No persistent memory of its own; purely a structured context-injection contract. DSPy's optimisation loop (MIPRO, BootstrapFewShot) treats historical turns as trainable few-shot structure.","cited_cells":44,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/dspy-history--dspy-ai"},{"id":"duolingo-max--blog-duolingo-com","name":"Duolingo Max","url":"https://blog.duolingo.com/duolingo-max/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"SRS state + GPT-4 roleplay continuity","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free + paid","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"relational","retrieval":"injection","persistence":"lifelong","update":"overwrite","unit":"profile|fact","governance":"inspectable","conflict":"overwrite","description":"Adds GPT-4-powered Roleplay (character conversations, with cross-call memory for \"Lily\") and Explain My Answer on top of Duolingo's existing spaced-repetition engine, which tracks per-word and per-skill mastery state across all sessions.","cited_cells":34,"third_party_cells":12,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/duolingo-max--blog-duolingo-com"},{"id":"elicit--elicit-com","name":"Elicit","url":"https://elicit.com/","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Session-persistent claim-to-sentence provenance","license":"","latest_release":"","deployment":"Managed-only; enterprise single-tenancy available","pricing":"Basic free; Pro $49/user/mo; Scale $169/user/mo; Enterprise custom (annual billing; page showed conflicting variants incl. Plus $11 / Pro $39 / Scale $89)","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"relational","retrieval":"exact-match|similarity","persistence":"cross-session","update":"extraction","unit":"document|fact","governance":"auditable","conflict":"append","description":"Systematic literature review + data extraction + meta-analysis. Every extracted claim anchored to its exact source sentence; Notebook persists paper collections across projects. Research Agents (Dec 2025) carry evidence forward across review stages.","cited_cells":28,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/elicit--elicit-com"},{"id":"elsa-speak--elsaspeak-com","name":"ELSA Speak","url":"https://elsaspeak.com/en/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Evolving phoneme-level pronunciation profile","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free + paid individual plans; Business (50+ learners) and Schools (21+) custom pricing","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"relational|vector","retrieval":"injection|similarity","persistence":"lifelong","update":"extraction|overwrite","unit":"profile|fact","governance":"inspectable","conflict":"overwrite","description":"Persistent pronunciation profile per user, tracking which sounds, words, and patterns the learner consistently mispronounces. Self-evolving model re-evaluates and updates the profile as the learner improves. Competitive moat is a proprietary non-native accent training dataset, not a general LLM.","cited_cells":36,"third_party_cells":10,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/elsa-speak--elsaspeak-com"},{"id":"engram--gh-gentleman-programming-engram","name":"engram","url":"https://github.com/Gentleman-Programming/engram","section":"Claude Code memory mechanisms","tier":2,"status":"active","last_verified_at":"2026-10-23","type":"Static Go binary, agent-agnostic","license":"MIT","latest_release":"v3.5.2 (2026-10-21)","deployment":"local — single Go binary with SQLite/FTS5; MCP stdio, works with any MCP agent","pricing":"free / open source (MIT)","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"Gentleman-Programming/engram — ~6.7k stars, last commit 2026-10","storage":"relational","retrieval":"exact-match","persistence":"long-term","update":"agent-controlled","unit":"fact|document","governance":"user-controllable","conflict":"append","description":"SQLite + FTS5; agent decides what to save (not firehose capture). Single Go binary, no CGO, no Node/Python/Docker dependencies. HTTP API + MCP + CLI + TUI; optional cloud sync with local-source-of-truth.","cited_cells":42,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/engram--gh-gentleman-programming-engram"},{"id":"enkrypt-ai--enkryptai-com","name":"Enkrypt AI","url":"https://www.enkryptai.com/","section":"Memory governance, privacy & safety","tier":1,"status":"acquired","last_verified_at":"2026-10-29","type":"Three-point RAG / memory pipeline guardrails","license":"","latest_release":"","deployment":"SaaS for self-serve tiers; Enterprise offers VPC / on-prem deployment","pricing":"Explore free; Launch $149/mo ($134/mo annual); Scale $1,499/mo ($1,349/mo annual); Enterprise custom","mcp_support":"yes — MCP server scanning (Launch tier) and MCP Gateway + Registry (Scale/Enterprise)","a2a_support":"no data — searched enkryptai.com, enkryptai.com/platform/capabilities/ai-risk-removal, docs.enkryptai.com; A2A protocol not advertised","backend_storage":"custom","llm_lock":"","github":"","storage":"n/a","retrieval":"injection","persistence":"session","update":"read-only","unit":"policy","governance":"auditable","conflict":"n/a","description":"Applies guardrails at three points: (1) before write to vector DB, (2) before query reaches embedding model, (3) before response. Detects malicious instructions in stored memory before retrieval; scans for PII/PHI in/out. Text + image + voice modalities. Acquired by Anaconda (2026-08-04); Enkrypt AI is now sold as Anaconda's AI security and guardrails product.","cited_cells":42,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/enkrypt-ai--enkryptai-com"},{"id":"era-computer--techcrunch-com","name":"Era Computer","url":"https://era.world/","section":"Voice-first / wearable AI memory","tier":5,"status":"active","last_verified_at":"2026-10-29","type":"User-sovereign pluggable memory layer for wearable OS","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"","storage":"proprietary","retrieval":"agentic","persistence":"long-term","update":"agent-controlled","unit":"file","governance":"user-controllable","conflict":"none","description":"Software platform / OS layer for AI wearables (glasses, rings, pendants, speakers) that abstracts hardware from AI orchestration. Memory and model providers are pluggable and user-controlled rather than cloud-locked; supports 130+ LLMs from 14+ providers. Founded 2025 by ex-Humane team.","cited_cells":19,"third_party_cells":17,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/era-computer--techcrunch-com"},{"id":"eve-shared-firm-knowledge-auditor--eve-legal","name":"Eve — Shared Firm Knowledge + Auditor","url":"https://www.eve.legal/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Firm-level store + nightly autonomous case auditor","license":"","latest_release":"","deployment":"searched not found","pricing":"No public pricing; demo / sales-led","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|graph","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction|agent-controlled","unit":"document|fact","governance":"auditable","conflict":"llm-arbitrate","description":"Plaintiff-law AI operating layer (EveOS) spanning intake, case work, research and firm analytics; its Auditor module reviews open matters for missed case value. The firm-wide knowledge store is no longer a named module on the homepage.","cited_cells":31,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/eve-shared-firm-knowledge-auditor--eve-legal"},{"id":"fellou--fellou-ai","name":"Fellou","url":"https://fellou.ai/","section":"Browser-agent memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Spatial agentic browser + Agentic Memory","license":"","latest_release":"","deployment":"Managed-only (browser app)","pricing":"Free (1,000-Spark bonus); Plus $19/mo; Pro $39.90/mo; Ultra $199.90/mo (annual ~20% off)","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"proprietary","retrieval":"agentic","persistence":"long-term","update":"agent-controlled","unit":"trajectory","governance":"opaque","conflict":"none","description":"Silicon Valley. Spatial agentic browser with explicit Agentic Memory layer learning from history and notes for instant recall of past research sessions. Operates on logged-in accounts (Quora, X, etc.); runs tasks while user is idle.","cited_cells":42,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/fellou--fellou-ai"},{"id":"figure-helix--figure-ai","name":"Figure Helix","url":"https://www.figure.ai/helix","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Temporal vision memory + tactile buffer","license":"","latest_release":"","deployment":"Managed hardware deployment (robots at partner sites)","pricing":"searched not found","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|kv","retrieval":"attention|cache-lookup","persistence":"session","update":"overwrite","unit":"episode|kv-token","governance":"opaque","conflict":"overwrite","description":"Humanoid VLA model: a 7B VLM (System 2, 7–9 Hz) continuously updates a shared latent vector of behavioral intent that an 80M visuomotor policy (System 1, 200 Hz) consumes alongside the latest observation; runs onboard Figure robots.","cited_cells":30,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/figure-helix--figure-ai"},{"id":"filevine-lois--filevine-com","name":"Filevine LOIS","url":"https://www.filevine.com/news/filevine-launches-the-only-embedded-ai-legal-assistant-that-lets-legal-professionals-chat-with-their-case/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Whole-matter context as live LLM input","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free LOIS Explore tier (research/drafting, no card) + 90-day free trial; paid/enterprise pricing via sales","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"relational|vector","retrieval":"injection|similarity","persistence":"long-term","update":"overwrite","unit":"document","governance":"auditable","conflict":"document-version","description":"Embedded in case-management platform. Holds full scope of every matter — deadlines, billing, discovery, communications, negotiation history — as live context for every query. Case data is the context, not a separate retrieval corpus.","cited_cells":35,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/filevine-lois--filevine-com"},{"id":"flare--arxiv-2305-06983","name":"FLARE","url":"https://arxiv.org/abs/2305.06983","section":"Retrieval-as-memory hybrids","tier":3,"status":"research-code","last_verified_at":"2026-10-23","type":"Confidence-triggered forward retrieval","license":"MIT","latest_release":"n/a","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"jzbjyb/FLARE — ~668 stars","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Forward-Looking Active Retrieval. Uses model's low-confidence token predictions as signal to anticipate future information needs and retrieve proactively. Retrieves only when token probability falls below threshold; generator stays frozen.","cited_cells":25,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/flare--arxiv-2305-06983"},{"id":"flowise-memory--docs-flowiseai-com","name":"Flowise Memory","url":"https://docs.flowiseai.com/integrations/langchain/memory","section":"Framework-embedded memory","tier":1,"status":"archived","last_verified_at":"2026-10-29","type":"Buffer + Buffer-Window + Conversation-Summary nodes","license":"Apache-2.0","latest_release":"flowise@3.1.4 (2026-07-29) — final","deployment":"Both (OSS self-hosted; Flowise Cloud managed; air-gapped deployment supported)","pricing":"discontinued — product sunset (EOL 2026-08-31); OSS code remains forkable under Apache-2.0","mcp_support":"via official adapter — Flowise MCP node","a2a_support":"searched not found","backend_storage":"pluggable","llm_lock":"any (BYO via LangChain LLMs)","github":"FlowiseAI/Flowise — ~55.4k stars, archived 2026-08-13 (sunset plan had targeted 2026-08-10)","storage":"vector|kv","retrieval":"similarity|injection","persistence":"session","update":"append-only|consolidation","unit":"episode|summary","governance":"inspectable","conflict":"append-only","description":"Discontinued: Flowise (acquired by Workday, Aug 2025) was sunset in 2026 — code freeze 2026-07-29, GitHub repo archived 2026-08-13, end of life 2026-08-31. Was a visual LangChain canvas with LangChain-native memory nodes (Buffer, Buffer Window, Conversation Summary Buffer) backed by an internal chat_message DB table; Buffer Window kept the last K interactions, Summary Buffer compressed older turns while keeping recent ones verbatim.","cited_cells":51,"third_party_cells":9,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/flowise-memory--docs-flowiseai-com"},{"id":"friend--friend-com","name":"Friend","url":"https://friend.com","section":"Voice-first / wearable AI memory","tier":1,"status":"active","last_verified_at":"2026-10-28","type":"Always-on AI companion pendant","license":"","latest_release":"","deployment":"Managed-only","pricing":"Hardware-bundled","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"lifelong","update":"append-only","unit":"episode","governance":"opaque","conflict":"append","description":"Pendant necklace (~2 inches). Always-on listening feeds a persistent AI companion persona. Companionship-framed rather than productivity / recall. Cloud LLM backend; no detailed encryption claims.","cited_cells":36,"third_party_cells":9,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/friend--friend-com"},{"id":"futurehouse--futurehouse-org","name":"FutureHouse","url":"https://www.futurehouse.org/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Cross-experiment + organism-scale factbase","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"extraction","unit":"fact|document","governance":"opaque","conflict":"llm-arbitrate","description":"Full autonomous-scientist loop. Owl agent surfaces prior work so downstream agents don't re-derive known results. WikiCrow produced cited factsheets for 15,616 previously undocumented human genes; Phoenix uses accumulated experiment context to plan chemistry work. FutureHouse is a 501(c)(3) non-profit research lab (founded 2023 by Sam Rodriques and Andrew White) whose flagship program is now an AI-for-Science postdoctoral fellowship; its tools are commercialized by Edison Scientific, a for-profit spun out of FutureHouse in 2025, and platform.futurehouse.org now redirects to Edison's platform (Kosmos agents for biopharma R&D). Recent FutureHouse research includes Robin, an end-to-end multi-agent discovery system (2026-05).","cited_cells":28,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/futurehouse--futurehouse-org"},{"id":"galileo-galileo-ai--galileo-ai","name":"Galileo (galileo.ai) (now Splunk Agent Observability)","url":"https://www.splunk.com/en_us/about-splunk/acquisitions/galileo.html","section":"Memory observability & monitoring","tier":1,"status":"acquired","last_verified_at":"2026-10-28","type":"Real-time intent / belief drift detection","license":"","latest_release":"","deployment":"SaaS cloud; Enterprise: VPC or on-prem","pricing":"Standalone pricing retired — galileo.ai/pricing now redirects to Splunk pricing; sold as Splunk Agent Observability","mcp_support":"Yes — Galileo MCP Server documented; integrates with traces, prompts, datasets","a2a_support":"Yes — A2A is listed under Integrations Overview (alongside CrewAI, Google ADK, LangChain, etc.)","backend_storage":"custom","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"cross-session","update":"append-only","unit":"episode","governance":"auditable","conflict":"n/a","description":"Acquired by Cisco (announced 2026-04-09, closed by 2026-05); now Splunk Agent Observability. Treats memory as first-class in multi-agent tracing. Luna-2 SLMs (3B / 8B) scan every interaction for intent drift and belief drift; 20+ checks at sub-200ms latency. Catches when agent A's view of the world splits from teammate B's. OpenTelemetry-compatible.","cited_cells":45,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/galileo-galileo-ai--galileo-ai"},{"id":"gbrain-garry-tan--gh-garrytan-gbrain","name":"gbrain (Garry Tan)","url":"https://github.com/garrytan/gbrain","section":"Dedicated memory layers","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"MIT","latest_release":"last commit 2026-10-22","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/garrytan/gbrain (31075★)","storage":"file|relational|vector","retrieval":"hybrid-rerank|graph-traversal|injection","persistence":"long-term","update":"agent-controlled|extraction|user-edit","unit":"document|fact|entity","governance":"auditable|editable","conflict":"human-arbitrate","description":"Open-source agent 'brain' storing explicit sourced facts with corrections and withdrawals, shared across agents; built as the memory for OpenClaw/Hermes agents. author is YC CEO; listed #6 in TeleAI Awesome-Agent-Memory","cited_cells":16,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/gbrain-garry-tan--gh-garrytan-gbrain"},{"id":"gemini-cli-memory-save-memory-gemini-md--geminicli-com","name":"Gemini CLI memory (save_memory / GEMINI.md)","url":"https://geminicli.com/docs/tools/memory","section":"Coding-agent memory","tier":3,"status":"active","last_verified_at":"2026-10-29","type":"Product","license":"Apache-2.0","latest_release":"last commit 2026-10-21","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/google-gemini/gemini-cli (107516★)","storage":"file","retrieval":"injection","persistence":"cross-session|project-bound","update":"agent-controlled|overwrite","unit":"fact|file","governance":"editable|plaintext","conflict":"agent","description":"Google's open-source Gemini CLI persists facts by editing tiered Markdown memory files — repo GEMINI.md, a per-project private memory folder and global ~/.gemini/GEMINI.md — that load into hierarchical context each session. Traction: Google first-party; Gemini CLI is one of the most-used terminal coding agents. On 2026-06-18 Google replaced Gemini CLI with Antigravity CLI for free, Google One and Google AI Pro/Ultra users; Gemini CLI stays supported for Gemini Code Assist Standard/Enterprise licenses and paid Gemini or Agent Platform API keys. As of 2026-10 the repo README describes Gemini CLI as maintained for enterprise Gemini Code Assist workflows and not tuned for the Gemini 4 model series.","cited_cells":17,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/gemini-cli-memory-save-memory-gemini-md--geminicli-com"},{"id":"generative-semantic-workspace-gsw--arxiv-2511-07587","name":"Generative Semantic Workspace (GSW)","url":"https://arxiv.org/abs/2511.07587","section":"Retrieval-as-memory hybrids","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Episodic memory for RAG","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"roychowdhuryresearch/gsw-memory — ~25 stars","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"read-only","unit":"episode","governance":"inspectable","conflict":"llm-arbitrate","description":"Replaces chunk-retrieval with neuro-inspired generative memory. Operator maps observations to intermediate semantic structures; Reconciler integrates them into a persistent workspace with temporal/spatial/logical coherence. Targets 100k–1M-token corpora.","cited_cells":25,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/generative-semantic-workspace-gsw--arxiv-2511-07587"},{"id":"github-copilot-custom-instructions--gh-copilot-customizing-copilot","name":"GitHub Copilot custom instructions","url":"https://docs.github.com/copilot/customizing-copilot/adding-custom-instructions-for-github-copilot","section":"File-backed / editor paradigms","tier":1,"status":"active","last_verified_at":"2026-10-22","type":".github/copilot-instructions.md + path-scoped","license":"","latest_release":"","deployment":"not applicable — GitHub-hosted; no separate deployment","pricing":"not applicable — file pattern included in GitHub Copilot subscription ($10/mo individual; $19/mo business)","mcp_support":"","a2a_support":"","backend_storage":"file (.github/copilot-instructions.md)","llm_lock":"","github":"not applicable — docs feature, no GitHub repo","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"manual","description":".github/copilot-instructions.md (repo-wide) plus .github/instructions/*.instructions.md path-scoped via applyTo glob (optional excludeAgent); Copilot agents also honor AGENTS.md (nearest in tree) or a root CLAUDE.md/GEMINI.md.","cited_cells":33,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/github-copilot-custom-instructions--gh-copilot-customizing-copilot"},{"id":"github-copilot-memory-agentic-memory--gh-en-copilot","name":"GitHub Copilot Memory (agentic memory)","url":"https://docs.github.com/en/copilot/concepts/agents/copilot-memory","section":"Coding-agent memory","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Product","license":"proprietary","latest_release":"searched not found","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"searched not found","storage":"proprietary","retrieval":"agentic","persistence":"cross-session","update":"agent-controlled|evict-oldest","unit":"fact|profile","governance":"user-controllable|inspectable","conflict":"grounded-ranking|llm-arbitrate","description":"Repository-scoped memories that Copilot coding agent, code review and CLI write automatically, share across features, validate against the current code before use, and expire after 28 days; owners can review/delete them. Traction: Shipped to all paid GitHub Copilot plans; default-on for Pro/Pro+","cited_cells":13,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/github-copilot-memory-agentic-memory--gh-en-copilot"},{"id":"google-adk-memory--google-github-io","name":"Google ADK Memory","url":"https://google.github.io/adk-docs/sessions/memory/","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"SessionService tiers + VertexAIMemoryBank","license":"Apache-2.0","latest_release":"v2.6.3 (2026-08-07)","deployment":"Both (ADK OSS self-hosted; one-command deploy to Google Cloud Agent Engine / Cloud Run / GKE)","pricing":"ADK OSS free; Vertex AI Memory Bank and Agent Engine: consumption-based pay-per-use cloud pricing (serverless)","mcp_support":"native (first-party) — ADK MCP integration","a2a_support":"supported (native) — ADK is A2A reference impl","backend_storage":"pluggable (Vertex AI Memory Bank)","llm_lock":"multi-vendor (Vertex/Anthropic/OpenAI via plugins)","github":"google/adk-python — ~21.1k stars, last commit 2026-08","storage":"kv|vector","retrieval":"injection|similarity","persistence":"cross-session","update":"overwrite|extraction","unit":"episode|fact","governance":"inspectable","conflict":"overwrite","description":"Google's Agent Development Kit. Session services: InMemory / Database / VertexAI. Long-term: InMemoryMemoryService or VertexAIMemoryBankService (uses Vertex AI Memory Bank Preview for persistent cross-session memory).","cited_cells":57,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/google-adk-memory--google-github-io"},{"id":"google-gemini-memory--blog-google","name":"Google Gemini Memory","url":"https://blog.google/products/gemini/temporary-chats-privacy-controls/","section":"Platform-provider memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Personal Context + Personal Intelligence","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free; Google AI Pro $20/mo; AI Ultra $100/mo and $200/mo tiers (US, May 2026)","mcp_support":"via official adapter — Gemini supports MCP via ADK / extensions","a2a_support":"supported — Google originated A2A","backend_storage":"custom (Google-managed)","llm_lock":"","github":"","storage":"kv","retrieval":"injection","persistence":"long-term","update":"extraction","unit":"fact|profile","governance":"user-controllable","conflict":"llm-arbitrate","description":"Gemini app learns from past chats via the Personal context setting (on by default, can be disabled; temporary chats excluded). Users can import memories and full chat history from ChatGPT, Claude and other assistants (18+, personal accounts, not EEA/CH/UK).","cited_cells":70,"third_party_cells":10,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/google-gemini-memory--blog-google"},{"id":"granola--granola-ai","name":"Granola","url":"https://www.granola.ai","section":"Voice-first / wearable AI memory","tier":1,"status":"active","last_verified_at":"2026-10-23","type":"Mac/iOS meeting notepad — system-audio tap","license":"","latest_release":"","deployment":"Managed-only","pricing":"Basic $0; Business $14/user/mo; Enterprise $35/user/mo","mcp_support":"yes — official Granola MCP server on all tiers (Basic: last 30 days of personal notes; Business/Enterprise: workspace notes, admin controls)","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"file","retrieval":"extraction-pull","persistence":"long-term","update":"append-only","unit":"episode|document","governance":"user-controllable","conflict":"append","description":"Mac/iOS app (no hardware). Captures meeting audio via system-audio tap (no bot joining); generates notes from your shorthand + audio. Local audio processing claim; cloud for summaries. Enterprise tier opts out of model training.","cited_cells":71,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/granola--granola-ai"},{"id":"graphiti-mcp-server-zep--getzep-com","name":"Graphiti MCP Server (Zep)","url":"https://www.getzep.com/product/knowledge-graph-mcp","section":"Claude Code memory mechanisms","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Temporal KG with bi-temporal edges","license":"Apache-2.0","latest_release":"v0.31.2 (2026-10-20)","deployment":"cloud (Zep managed) or self-hosted via Graphiti OSS + Neo4j/local graph DB","pricing":"free tier: 2500 messages/mo + 2.5MB graph data; paid: $1.25/1000 messages + $2.50/MB extra; Enterprise custom","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"Neo4j","llm_lock":"","github":"getzep/graphiti — ~31.9k stars, last commit 2026-10","storage":"graph","retrieval":"graph-traversal","persistence":"long-term","update":"append-only","unit":"episode|fact","governance":"auditable","conflict":"bi-temporal","description":"Tracks how facts change over time; maintains provenance; supports prescribed and learned ontology. Bi-temporal edges (when known vs when valid). Only server that distinguishes \"fact learned\" from \"fact true.\"","cited_cells":61,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/graphiti-mcp-server-zep--getzep-com"},{"id":"graphraft--arxiv-2504-05478","name":"GraphRAFT","url":"https://arxiv.org/abs/2504.05478","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"RAG fine-tuning over KGs","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"AlfredClemedtson/graphraft — ~7 stars","storage":"graph","retrieval":"graph-traversal","persistence":"parametric-permanent","update":"parametric-edit","unit":"document","governance":"opaque","conflict":"n/a","description":"Retrieval-augmented fine-tuning for knowledge graphs on graph databases.","cited_cells":23,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/graphraft--arxiv-2504-05478"},{"id":"graphrag-microsoft--microsoft-com","name":"GraphRAG (Microsoft)","url":"https://www.microsoft.com/en-us/research/project/graphrag/","section":"Retrieval-as-memory hybrids","tier":1,"status":"maintenance","last_verified_at":"2026-10-29","type":"Community summary + Leiden detection","license":"MIT","latest_release":"v3.3.0 (2026-10-08)","deployment":"searched not found","pricing":"searched not found","mcp_support":"not documented publicly","a2a_support":"not supported","backend_storage":"Parquet files / pluggable","llm_lock":"","github":"microsoft/graphrag — ~35.1k stars, last commit n/a (PyPI release 2026-10)","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"extraction","unit":"document|summary","governance":"inspectable","conflict":"n/a","description":"Modular graph-based RAG. Leiden community detection over knowledge graph; community summaries at multiple levels. DRIFT Search hybrid (Oct 2024). The GraphRAG repo is now largely in maintenance mode: no new features or PRs, only bug fixes and dependency/CVE updates.","cited_cells":54,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/graphrag-microsoft--microsoft-com"},{"id":"grok-memory-xai--maginative-com","name":"Grok Memory (xAI)","url":"https://x.com/grok/status/1912670182012801156","section":"Platform-provider memory","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Preference-distillation + user-deletable","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free; The X Subscription (Grok + Cursor + X): Plus $8/mo, Premium $30/mo, Super $100/mo, Ultra $200/mo (annual discounts; team seats on Premium/Super/Ultra); existing SuperGrok plans continue","mcp_support":"searched not found","a2a_support":"not supported","backend_storage":"custom (xAI-managed)","llm_lock":"","github":"","storage":"kv","retrieval":"injection","persistence":"long-term","update":"extraction","unit":"fact","governance":"user-controllable","conflict":"llm-arbitrate","description":"Extracts and persists facts from past conversations; applies them to future responses. Per-memory deletion UI; user can also clear all. Announced by xAI's official Grok account on 2025-04-17 ('Grok now remembers your conversations').","cited_cells":60,"third_party_cells":23,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/grok-memory-xai--maginative-com"},{"id":"gwm-gaussian-world-models--gaussian-world-model-github-io","name":"GWM (Gaussian World Models)","url":"https://gaussian-world-model.github.io/","section":"Vertical / domain-specific AI memory","tier":3,"status":"stale","last_verified_at":"2026-10-28","type":"3D Gaussian Splatting latent world model","license":"MIT","latest_release":"","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"Gaussian-World-Model/gaussianwm — ~143 stars, last commit 2025-10","storage":"parametric","retrieval":"parametric-recall","persistence":"session","update":"overwrite","unit":"scene-graph","governance":"inspectable","conflict":"overwrite","description":"Encodes scene into 3D Gaussian Splat representations via 3D VAE; latent Diffusion Transformer predicts future Gaussian Splat states conditioned on robot actions. 3D scene representation persists across time steps as structural scene memory rather than flat features.","cited_cells":23,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/gwm-gaussian-world-models--gaussian-world-model-github-io"},{"id":"harvey-memory--harvey-ai","name":"Harvey Memory","url":"https://www.harvey.ai/blog/memory-in-harvey","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"4-tier matter / firm / client-institution memory","license":"","latest_release":"","deployment":"Both","pricing":"Enterprise only","mcp_support":"yes — Harvey acts as MCP client via MCP Connector Library (vendors/firms plug tools in), with an MCP policy engine; own MCP server only planned (Dec 2025)","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|graph","retrieval":"similarity|graph-traversal|injection","persistence":"long-term","update":"extraction","unit":"episode|document|fact","governance":"auditable","conflict":"none","description":"Memory carries forward user-chosen context (matter details, precedent, preferences, approved practices) and reuses past threads scoped to a user, client, matter or firm process. Firm admins enable/disable it; ethical walls and audit controls promised. Announced Jan 2026 as co-development with customers.","cited_cells":60,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/harvey-memory--harvey-ai"},{"id":"haystack-memory-deepset--haystack-deepset-ai","name":"Haystack Memory (deepset)","url":"https://haystack.deepset.ai/","section":"Framework-embedded memory","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Pipeline-native ChatMessageStore","license":"Apache-2.0","latest_release":"v3.4.1 (2026-10-21)","deployment":"Both (OSS self-hosted; Haystack Enterprise Platform: SaaS cloud, VPC, or on-prem)","pricing":"OSS free (Apache-2.0); Haystack Enterprise Platform: custom licensing + runtime fees + optional expert services","mcp_support":"via official adapter — haystack-mcp-tools","a2a_support":"searched not found","backend_storage":"","llm_lock":"any (BYO via Generators)","github":"deepset-ai/haystack — ~26.2k stars, last commit 2026-10","storage":"kv","retrieval":"exact-match|injection","persistence":"session","update":"append-only","unit":"document","governance":"inspectable","conflict":"append","description":"Haystack 2.x: InMemoryChatMessageStore (experimental) + ChatMessageRetriever + ChatMessageWriter components in declarative pipelines. Legacy: ConversationSummaryMemory with LLM-compressed rolling summaries. Memory is an explicit pipeline component with swappable backends.","cited_cells":53,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/haystack-memory-deepset--haystack-deepset-ai"},{"id":"hebbia-matrix--hebbia-com","name":"Hebbia Matrix","url":"https://www.hebbia.com/","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Firm-specific configuration over multi-agent synthesis","license":"","latest_release":"","deployment":"searched not found","pricing":"not published — sales/demo only","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"cross-session","update":"extraction","unit":"document","governance":"auditable","conflict":"none","description":"Corporate / transactional due diligence. Parallel AI agents across large document sets with firm-specific configurations and review workflows. Configuration amounts to a shallow institutional memory layer; cross-matter persistent memory not documented.","cited_cells":29,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/hebbia-matrix--hebbia-com"},{"id":"heidi-health--heidihealth-com","name":"Heidi Health","url":"https://www.heidihealth.com/","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Clinician-voice / provider-side personalisation","license":"","latest_release":"","deployment":"searched not found","pricing":"searched not found","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|relational","retrieval":"similarity|injection","persistence":"cross-session","update":"extraction","unit":"fact|profile","governance":"auditable","conflict":"editor-in-the-loop","description":"Ambient documentation with clinician-profile learning. System learns each clinician's preferred note structure, phrasing patterns, symptom-outlining conventions — applies persistently across sessions. Provider-side personalisation rather than patient-longitudinal memory.","cited_cells":30,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/heidi-health--heidihealth-com"},{"id":"heptabase--heptabase-com","name":"Heptabase","url":"https://heptabase.com/","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Visual whiteboard + cards + AI chat with citations","license":"","latest_release":"","deployment":"Managed-only","pricing":"Pro $8.99/mo; Premium $17.99/mo; Premium+ $53.99/mo (yearly billing saves 25%)","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"custom (local SQLite + sync)","llm_lock":"","github":"","storage":"graph|file","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"overwrite","description":"Visual whiteboard canvas + nested cards + bidirectional links + journal. AI chat with cited sources inside whiteboards; supports OpenAI / Gemini / Anthropic backends.","cited_cells":37,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/heptabase--heptabase-com"},{"id":"hermes-agent-memory-nous-research--gh-nousresearch-hermes-agent","name":"Hermes Agent memory (Nous Research)","url":"https://github.com/NousResearch/hermes-agent","section":"Framework-embedded memory","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"MIT","latest_release":"last commit 2026-10-22","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/NousResearch/hermes-agent (256157★)","storage":"file|relational","retrieval":"injection|exact-match|agentic","persistence":"long-term","update":"agent-controlled|consolidation","unit":"fact|profile|skill","governance":"editable","conflict":"agent","description":"Nous Research's open-source self-improving agent keeps an agent-curated MEMORY.md/USER.md notebook injected at session start, full-text (FTS5) search over past sessions that returns raw messages without LLM summarisation, and optional Honcho user modelling. one of the most-starred agent projects of 2026.","cited_cells":16,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/hermes-agent-memory-nous-research--gh-nousresearch-hermes-agent"},{"id":"heyday--heyday-xyz","name":"Heyday","url":"https://www.heyday.xyz/","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"archived","last_verified_at":"2026-10-29","type":"Browser-based research memory","license":"","latest_release":"","deployment":"Managed-only (browser extension + cloud)","pricing":"$19/mo (monthly) or $179/yr (annual) — browser extension AI research assistant","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"append-only","unit":"document","governance":"user-controllable","conflict":"none","description":"AI assistant that builds a searchable memory from meetings (Zoom app), emails, documents, notes and web reading; answers questions, prepares client context and drafts content. Now marketed mainly to coaches and knowledge workers. As of 2026-10-29 heyday.xyz is a GoDaddy domain-parking page, so the product appears to be gone (no shutdown announcement found).","cited_cells":30,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/heyday--heyday-xyz"},{"id":"hiddenlayer-aisec-platform-2-0--hiddenlayer-com","name":"HiddenLayer AISec Platform 2.0","url":"https://www.hiddenlayer.com/platform","section":"Memory governance, privacy & safety","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Model Genealogy + AIBOM (supply-chain layer)","license":"","latest_release":"","deployment":"Agentless, model-agnostic; connectors for cloud, CI/CD, SIEM/SOAR, API gateways, MLOps (specific SaaS/on-prem options not stated)","pricing":"enterprise pricing; contact sales; no public pricing","mcp_support":"Yes — Agentic and MCP Security use-case (Use case 04); protects agents and MCP-based systems from prompt injection / unsafe tool use","a2a_support":"no data — searched hiddenlayer.com, hiddenlayer.com/platform, docs.hiddenlayer.com; A2A protocol not advertised","backend_storage":"custom","llm_lock":"","github":"","storage":"n/a","retrieval":"n/a","persistence":"long-term","update":"read-only","unit":"model","governance":"auditable","conflict":"n/a","description":"AI security platform with four modules: AI asset discovery, supply-chain/model integrity validation, continuous attack simulation, and runtime detection/response, including protection of agents and MCP-based systems.","cited_cells":41,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/hiddenlayer-aisec-platform-2-0--hiddenlayer-com"},{"id":"hindsight-mcp--hindsight-vectorize-io","name":"Hindsight MCP","url":"https://hindsight.vectorize.io/developer/mcp-server","section":"Claude Code memory mechanisms","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Four-network episodic memory","license":"MIT","latest_release":"v1.1.0 (2026-10-21)","deployment":"Both (self-hosted server; Hindsight Cloud managed)","pricing":"OSS self-hosted free; Hindsight Cloud usage-based with free starter credits","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"vectorize-io/hindsight — ~44.6k stars, last commit 2026-10","storage":"graph","retrieval":"graph-traversal","persistence":"long-term","update":"extraction","unit":"fact|episode","governance":"inspectable","conflict":"llm-arbitrate","description":"World (facts), Experiences (agent actions), Opinion (beliefs with confidence scores), Observation (mental models). Entity resolution (\"Alice\" and \"my coworker Alice\" same node). Mimics human recall topology.","cited_cells":35,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/hindsight-mcp--hindsight-vectorize-io"},{"id":"hindsight-vectorize--hindsight-vectorize-io","name":"Hindsight (Vectorize)","url":"https://hindsight.vectorize.io/","section":"Dedicated memory layers","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Vector + reflection / summarisation","license":"MIT","latest_release":"v1.1.0 (2026-10-21)","deployment":"Both — managed Hindsight Cloud; Enterprise single-tenant (Vectorize cloud, customer AWS/GCP/Azure, or on-prem); self-host via Docker, Helm or pip (Postgres + pgvector)","pricing":"OSS self-host free (MIT); Hindsight Cloud pay-as-you-go credits ($5 free credit on GitHub sign-up); Enterprise single-tenant custom","mcp_support":"native (first-party) — Hindsight MCP","a2a_support":"not documented publicly","backend_storage":"custom","llm_lock":"","github":"vectorize-io/hindsight — ~44.6k stars, last commit 2026-10","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"consolidation","unit":"episode|summary","governance":"opaque","conflict":"none","description":"Standalone memory service from Vectorize. Open source. Biomimetic four-network design (World, Bank, Observation, Opinion). Ships an MCP memory server.","cited_cells":62,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/hindsight-vectorize--hindsight-vectorize-io"},{"id":"hippocratic-ai-polaris--hippocraticai-com","name":"Hippocratic AI Polaris","url":"https://hippocraticai.com/polaris-3/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Longitudinal multi-call patient memory store","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity|injection","persistence":"long-term","update":"append-only","unit":"episode|fact","governance":"auditable","conflict":"constellation-of-agents","description":"Patient-conversation / chronic-care staffing model. HIPAA-compliant Memory Store holds non-EHR salient facts from prior calls, persists across engagements, explicitly not used for model training. Shifts from transactional call-handling to longitudinal patient relationship.","cited_cells":60,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/hippocratic-ai-polaris--hippocraticai-com"},{"id":"hipporag-2--arxiv-2502-14802","name":"HippoRAG 2","url":"https://arxiv.org/abs/2502.14802","section":"Retrieval-as-memory hybrids","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Hippocampus-inspired RAG v2","license":"MIT","latest_release":"n/a","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"OSU-NLP-Group/HippoRAG — ~4.0k stars, last commit n/a","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"accretion","unit":"fact","governance":"plaintext","conflict":"n/a","description":"OSU follow-on to HippoRAG — adds online updates, recall-precision tradeoff calibration, and multi-hop reasoning improvements.","cited_cells":12,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/hipporag-2--arxiv-2502-14802"},{"id":"hipporag-hipporag2--gh-osu-nlp-group-hipporag","name":"HippoRAG / HippoRAG2","url":"https://github.com/OSU-NLP-Group/HippoRAG","section":"Retrieval-as-memory hybrids","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Personalised PageRank over KG","license":"MIT","latest_release":"v2.0.0a4 (2025-06-24) — PyPI pre-release; last GitHub release v1.0.0 (2025-02-27)","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"OSU-NLP-Group/HippoRAG — ~4.0k stars","storage":"graph","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Hippocampus-inspired RAG. LLMs + knowledge graphs + Personalized PageRank to mimic neocortex/hippocampus roles. NeurIPS 2024.","cited_cells":31,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/hipporag-hipporag2--gh-osu-nlp-group-hipporag"},{"id":"hirag--arxiv-2503-10150","name":"HiRAG","url":"https://arxiv.org/abs/2503.10150","section":"Retrieval-as-memory hybrids","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Hierarchical KG indexing + retrieval","license":"MIT","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"hhy-huang/HiRAG — ~557 stars, last commit not confirmed","storage":"graph","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Multi-layer knowledge graph: higher layers hold coarse summary entities improving connectivity between similar lower-layer entities (HiIndex). HiRetrieval traverses across granularity levels.","cited_cells":24,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/hirag--arxiv-2503-10150"},{"id":"honcho-plastic-labs--honcho-dev","name":"Honcho (Plastic Labs)","url":"https://honcho.dev/","section":"Dedicated memory layers","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Product","license":"AGPL-3.0","latest_release":"last commit 2026-10-21","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/plastic-labs/honcho (7730★)","storage":"relational|vector","retrieval":"agentic|hybrid|injection","persistence":"long-term","update":"extraction|consolidation","unit":"profile|fact|summary","governance":"inspectable","conflict":"llm-arbitrate","description":"Memory library/service for stateful agents that reasons over stored messages to build 'peer' representations (theory-of-mind style user models) queryable via a dialectic API; used as Hermes Agent's and OpenClaw's user-modelling backend. integrated into Hermes Agent and OpenClaw; public evals at evals.honcho.dev","cited_cells":18,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/honcho-plastic-labs--honcho-dev"},{"id":"hyde--arxiv-2212-10496","name":"HyDE","url":"https://arxiv.org/abs/2212.10496","section":"Retrieval-as-memory hybrids","tier":3,"status":"research-code","last_verified_at":"2026-10-22","type":"Hypothetical document embeddings","license":"Unlicensed","latest_release":"no releases","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"texttron/hyde — ~582 stars, dormant","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"read-only","unit":"document","governance":"inspectable","conflict":"n/a","description":"LLM generates a hypothetical answer document; embeds that instead of the query. Closes the query-document semantic gap for retrieval. arxiv 2212.10496.","cited_cells":22,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/hyde--arxiv-2212-10496"},{"id":"hyperconsciousness-hc--gh-louis030195-hyperconsciousness","name":"Hyperconsciousness (hc)","url":"https://github.com/louis030195/hyperconsciousness","section":"Memory governance, privacy & safety","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"MIT","latest_release":"last commit 2026-10-21","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/louis030195/hyperconsciousness (8★)","storage":"append-only","retrieval":"exact-match","persistence":"long-term","update":"append-only","unit":"document|file","governance":"user-controllable|auditable","conflict":"append-only","description":"Developer-alpha encrypted, append-only knowledge store for humans and agents, with CLI/MCP access and scoped, expiring grants. Traction: Submitted via intake issue #181","cited_cells":16,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/hyperconsciousness-hc--gh-louis030195-hyperconsciousness"},{"id":"hyperspell--ycombinator-com","name":"Hyperspell","url":"https://www.ycombinator.com/companies/hyperspell","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Workplace agent memory connectors","license":"","latest_release":"","deployment":"Cloud API + SDK (no self-hosted option found)","pricing":"not public — book a demo","mcp_support":"unclear — official hyperspell-mcp repo is archived; current integration via SDKs/CLI into Claude Code, Codex, Cursor","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity|extraction-pull","persistence":"long-term","update":"extraction","unit":"document|fact","governance":"opaque","conflict":"none","description":"YC F25 'company brain': connects 50+ sources (Slack, Gmail, Notion, Drive...) into a permission-aware context graph served to agents via API/SDK, returning structured results or LLM-ready summaries; learns from queries over time.","cited_cells":35,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/hyperspell--ycombinator-com"},{"id":"inngest-agentkit-network-state--agentkit-inngest-com","name":"Inngest AgentKit — Network State","url":"https://agentkit.inngest.com/concepts/memory","section":"Framework-embedded memory","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Durable-network-state / event-driven","license":"Apache-2.0","latest_release":"","deployment":"Both (deploy AgentKit to preferred cloud provider; Inngest platform manages durable state)","pricing":"Free tier; Pro; Enterprise custom; exact pricing not published — contact for enterprise","mcp_support":"yes — MCP servers usable as agent tools (documented)","a2a_support":"searched not found","backend_storage":"","llm_lock":"any (BYO)","github":"inngest/agent-kit — ~926 stars, last commit n/a","storage":"kv","retrieval":"exact-match","persistence":"session","update":"overwrite","unit":"kv-token","governance":"inspectable","conflict":"overwrite","description":"Memory as Network State — shared key-value store owned by the multi-agent network, readable by router / agents / tools. Persisted durably via Inngest infrastructure. Memory writes decoupled from response latency: agent responds immediately, background Inngest function handles DB write durably with retries.","cited_cells":48,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/inngest-agentkit-network-state--agentkit-inngest-com"},{"id":"intercom-fin--fin-ai","name":"Intercom Fin","url":"https://fin.ai","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"CRM attributes + roadmap lifecycle memory","license":"","latest_release":"","deployment":"Managed-only","pricing":"$0.99 per outcome (50/mo minimum on any-helpdesk plan); +$29/seat/mo on Intercom helpdesk; Pro add-on from $100/mo; 14-day free trial","mcp_support":"yes — Fin acts as MCP client (prebuilt Linear/Stripe/Shopify connectors + custom remote MCP servers); Intercom also hosts a remote MCP server at mcp.intercom.com","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"relational|vector","retrieval":"injection|similarity","persistence":"cross-session","update":"overwrite","unit":"fact|episode","governance":"inspectable","conflict":"none","description":"CX support. Currently CRM-backed customer attributes + conversation history lookup. Lifecycle-persistent memory (\"priorities learned day one considered in year 10\") on roadmap as of Pioneer 2025 — not yet shipped.","cited_cells":33,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/intercom-fin--fin-ai"},{"id":"interloom--interloom-com","name":"Interloom","url":"https://interloom.com/","section":"Dedicated memory layers","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Enterprise tacit-knowledge memory","license":"","latest_release":"","deployment":"Managed multi-tenant SaaS (AWS); private single-tenant VPC (Azure); private VPC with bring-your-own-model","pricing":"No seat fees (unlimited users); usage-based per agent interaction — prepaid volume packages or pay-as-you-go","mcp_support":"not documented publicly","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"","storage":"proprietary","retrieval":"extraction-pull","persistence":"long-term","update":"extraction","unit":"skill","governance":"auditable","conflict":"none","description":"Back-office operations platform where agents and domain experts triage and resolve cases; each resolved case feeds a context graph, and a 'MemoryRank' component reuses prior successful cases to route new work.","cited_cells":41,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/interloom--interloom-com"},{"id":"inworld-ai--inworld-ai","name":"Inworld AI","url":"https://inworld.ai","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Long-term character memory + Character Brain","license":"","latest_release":"","deployment":"Both","pricing":"Usage-based: TTS-2 $25/1M chars on demand ($10 Enterprise); TTS-2 Flash $15/1M; STT-1 $0.10/audio-hour; open LLMs from $0.07/1M input tokens","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|graph","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction|consolidation","unit":"episode|fact","governance":"opaque","conflict":"llm-arbitrate","description":"Formerly an NPC character engine with named Long-Term Memory; now sells realtime voice models (TTS-2, STT-1, speech-to-speech) and OpenAI-compatible open-model inference. Character/NPC memory is no longer the headline product.","cited_cells":38,"third_party_cells":10,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/inworld-ai--inworld-ai"},{"id":"iris-ai--iris-ai","name":"Iris.ai","url":"https://iris.ai/","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Researcher Workspace + persistent extraction","license":"","latest_release":"","deployment":"Managed-only","pricing":"not public — request a demo","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"cross-session","update":"extraction","unit":"document","governance":"inspectable","conflict":"append","description":"Enterprise R&D document-intelligence platform, now sold as three products (Axion data prep, Neuralith enterprise knowledge engine, RSpace R&D intelligence) under an 'AI harness' that controls what models read and traces each answer step.","cited_cells":25,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/iris-ai--iris-ai"},{"id":"iter-retgen--arxiv-2305-15294","name":"ITER-RETGEN","url":"https://arxiv.org/abs/2305.15294","section":"Retrieval-as-memory hybrids","tier":3,"status":"research-code","last_verified_at":"2026-10-23","type":"Output-conditioned iterative retrieval","license":"not applicable — no official code","latest_release":"n/a","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"not applicable — no official code found (arXiv listing gives no code link)","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Uses model's generated output as a rich context signal for the next retrieval round rather than just the original query. Processes all retrieved knowledge as a whole, preserving generation flexibility.","cited_cells":23,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/iter-retgen--arxiv-2305-15294"},{"id":"k-health--khealth-com","name":"K Health","url":"https://khealth.com/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"EMR-filtered two-agent pipeline","license":"","latest_release":"","deployment":"searched not found","pricing":"searched not found","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"relational|vector","retrieval":"similarity|injection","persistence":"long-term","update":"extraction","unit":"episode|fact","governance":"auditable","conflict":"none","description":"Primary care / virtual physician intake. Two-agent pipeline: one extracts the relevant EMR subset, second answers grounded to that filtered history. Architecture explicitly diverges by patient condition (diabetic vs heart-failure get different responses to same intake).","cited_cells":28,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/k-health--khealth-com"},{"id":"kag--arxiv-2409-13731","name":"KAG","url":"https://arxiv.org/abs/2409.13731","section":"Retrieval-as-memory hybrids","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"KG-grounded hybrid reasoning","license":"Apache-2.0","latest_release":"v0.8.0 2025-06-28","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"OpenSPG/KAG — ~9.0k stars","storage":"graph","retrieval":"graph-traversal|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Knowledge-Augmented Generation. LLM-friendly KG representation + mutual indexing between KGs and raw chunks + logical-form-guided hybrid reasoning. Handles numerical / temporal / multi-hop chains that vector similarity misses.","cited_cells":28,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/kag--arxiv-2409-13731"},{"id":"khanmigo--khanmigo-ai","name":"Khanmigo","url":"https://www.khanmigo.ai/","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Chat history + Khan Academy mastery state","license":"","latest_release":"","deployment":"Managed-only","pricing":"Teachers free; learners/families $4/mo or $44/yr; district pricing on request","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"relational","retrieval":"injection","persistence":"cross-session","update":"append-only","unit":"episode","governance":"auditable","conflict":"append","description":"Stores full chat transcripts per student (visible to learner and teacher via dashboard). Layers over Khan Academy's existing mastery graph (exercise completion, skill progress) for cross-session topic awareness; cross-session memory is derived from structured mastery state rather than free-text recall.","cited_cells":35,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/khanmigo--khanmigo-ai"},{"id":"krutrim-kruti-ola--ai-labs-olakrutrim-com","name":"Krutrim Kruti (Ola)","url":"https://www.olakrutrim.com/kruti","section":"Dedicated memory layers","tier":2,"status":"archived","last_verified_at":"2026-10-29","type":"Cross-session multimodal agentic memory","license":"","latest_release":"","deployment":"Cloud-hosted on Indian servers (Krutrim Cloud; sovereign India-first)","pricing":"n/a — service offline since Apr 2026, vendor page says under maintenance (was free beta)","mcp_support":"announced (enterprise MCP and agent-to-agent protocols) at Jun 2025 launch; service now offline","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"cross-session","update":"extraction","unit":"episode|fact","governance":"opaque","conflict":"llm-arbitrate","description":"Offline since Apr 2026 (pulled from app stores and the Ola platform per press reports); Ola Krutrim's own Kruti page says (checked 2026-10-29) that Kruti AI is under maintenance and will be back shortly, with no official shutdown statement. Was Ola Krutrim's agentic multimodal assistant (beta Jun 2025; booking, food ordering, bill pay; MCP/A2A for enterprises).","cited_cells":34,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/krutrim-kruti-ola--ai-labs-olakrutrim-com"},{"id":"labgenius--labgenius-app","name":"LabGenius","url":"https://labgeniustx.com/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Autonomous experiment memory (EVA platform)","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"proprietary","retrieval":"agentic","persistence":"long-term","update":"extraction","unit":"episode|fact","governance":"opaque","conflict":"n/a","description":"Therapeutic-antibody discovery in a closed Design-Build-Test-Learn loop. EVA accumulates structured outcome data from each robotic experiment cycle and feeds it back to generative ML models directing the next design round. Genuine experiment-memory-as-training-signal.","cited_cells":30,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/labgenius--labgenius-app"},{"id":"lakera-guard-lakera-red--lakera-ai","name":"Lakera Guard / Lakera Red","url":"https://www.lakera.ai/ai-agent-security","section":"Memory governance, privacy & safety","tier":1,"status":"acquired","last_verified_at":"2026-10-29","type":"Real-time prompt-injection + poisoning screen","license":"","latest_release":"","deployment":"REST API (SaaS cloud) or self-hosted Enterprise deployment","pricing":"Developer: free (10K API calls/mo); Pro: contact sales (higher volume + advanced detection)","mcp_support":"no data — MCP referenced in Lakera blog as a threat-research topic only; no native MCP integration advertised (searched lakera.ai, docs.lakera.ai, lakera.ai/blog)","a2a_support":"no data — searched lakera.ai/lakera-guard, docs.lakera.ai, docs.lakera.ai/lakera-guard/integrations; A2A protocol not advertised","backend_storage":"custom","llm_lock":"","github":"","storage":"proprietary","retrieval":"injection","persistence":"session","update":"read-only","unit":"attack","governance":"auditable","conflict":"n/a","description":"Acquired by Check Point Software (completed by 2025-11-11); being folded into Check Point's AI security platform. Screens every prompt, response, and retrieved document for indirect prompt injection — primary vector for memory poisoning. Treats memory as untrusted: anything written to or read from memory is adversarial until proven clean. Lakera Red provides \"Agent Breaker\" gamified red-teaming.","cited_cells":43,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/lakera-guard-lakera-red--lakera-ai"},{"id":"langflow-memory--docs-langflow-org","name":"Langflow Memory","url":"https://docs.langflow.org/memory","section":"Framework-embedded memory","tier":1,"status":"acquired","last_verified_at":"2026-10-22","type":"SQLite buffer + Mem0/Redis integrations","license":"MIT","latest_release":"v1.11.4 (2026-08-19)","deployment":"OSS self-hosted (all major clouds); DataStax cloud deprecated; IBM acquisition path unclear","pricing":"OSS free (MIT); DataStax Langflow cloud was free tier + paid","mcp_support":"native (first-party) — Langflow MCP server","a2a_support":"searched not found","backend_storage":"pluggable","llm_lock":"any (BYO via LangChain)","github":"langflow-ai/langflow — ~156k stars, last commit 2026-08","storage":"relational|vector","retrieval":"injection|similarity","persistence":"cross-session","update":"append-only","unit":"episode","governance":"inspectable","conflict":"append-only","description":"Agent component has built-in chat memory enabled by default, keyed by session ID. Custom session IDs isolate per-user history. Bundled Mem0 + Redis integrations for external memory layers.","cited_cells":58,"third_party_cells":12,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/langflow-memory--docs-langflow-org"},{"id":"langfuse--langfuse-com","name":"Langfuse","url":"https://langfuse.com/","section":"Memory observability & monitoring","tier":1,"status":"acquired","last_verified_at":"2026-10-29","type":"Memory ops as named spans + Agent Graphs","license":"MIT (core; ee/ directories under separate enterprise license)","latest_release":"v4.27.0 (Sep 1; page shows no year, presumably 2026)","deployment":"SaaS cloud (Berlin + SF) or self-hosted Docker/K8s; ClickHouse native integration post-acquisition","pricing":"free: 50K observations/mo; Pro: $59/mo; Team: $199/mo; Enterprise: custom","mcp_support":"via official adapter — langfuse-mcp","a2a_support":"no data — searched langfuse.com, langfuse.com/docs, langfuse.com/integrations; A2A protocol not advertised (extensive framework integrations, MCP server present, OTel-native)","backend_storage":"Postgres + ClickHouse","llm_lock":"","github":"langfuse/langfuse — ~36.1k stars, last release 2026-09","storage":"relational","retrieval":"exact-match","persistence":"long-term","update":"append-only","unit":"episode","governance":"auditable","conflict":"n/a","description":"Acquired by ClickHouse (2026-01-16); product and MIT-licensed repo continue under the Langfuse name. Memory module reads/writes captured as named spans. Trace Log View concatenates every agent step including memory ops. Agent Graphs (GA 2025) infer graph structure from observation nesting; session-level replay tracks how memory state evolves.","cited_cells":51,"third_party_cells":8,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/langfuse--langfuse-com"},{"id":"langgraph-persistence--docs-langchain-com","name":"LangGraph Persistence","url":"https://docs.langchain.com/oss/python/langgraph/persistence","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Checkpoints + Store System","license":"MIT","latest_release":"langgraph v1.2.16 (2026-10-16)","deployment":"Both (LangGraph OSS self-hosted; LangSmith Deployment cloud managed or self-hosted)","pricing":"OSS free (MIT); LangSmith Developer $0 (5k traces/mo, 1 seat); Plus $39/seat/mo (1 free small serverless deployment); deployments metered at $1/LSU by compute+memory; Enterprise custom","mcp_support":"yes — Agent Server exposes deployed graphs as MCP tools at /mcp (streamable HTTP); MCP client via langchain[mcp] adapter","a2a_support":"supported — LangGraph A2A wrappers","backend_storage":"pluggable (Postgres, SQLite, Redis)","llm_lock":"any (BYO)","github":"langchain-ai/langgraph — ~43k stars, last commit 2026-10","storage":"relational|kv","retrieval":"exact-match|injection","persistence":"cross-session","update":"overwrite|append-only","unit":"episode|fact","governance":"inspectable","conflict":"last-write-wins","description":"Distinct from LangMem. Built-in checkpointer saves graph state per super-step (short-term, thread-scoped). Store System adds long-term hierarchical key-value memory across threads with optional vector search + TTL. Postgres / Mongo / Redis stores for production.","cited_cells":61,"third_party_cells":20,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/langgraph-persistence--docs-langchain-com"},{"id":"langmem-langchain--langchain-ai-github-io","name":"LangMem (LangChain)","url":"https://langchain-ai.github.io/langmem/","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-23","type":"Episodic / semantic / procedural","license":"MIT","latest_release":"v0.0.30 (2025-10-27) — PyPI","deployment":"Both (LangSmith cloud managed + self-hosted Kubernetes; LangGraph Platform cloud or self-hosted)","pricing":"OSS library free; LangSmith (observability/platform) Free 5k traces/mo; Plus $39/mo; Enterprise custom.","mcp_support":"via official adapter — LangChain MCP adapters","a2a_support":"supported (LangChain A2A integration)","backend_storage":"pluggable","llm_lock":"any (BYO)","github":"langchain-ai/langmem — ~1.6k stars, 147 commits","storage":"vector|kv","retrieval":"similarity|injection","persistence":"long-term","update":"extraction","unit":"episode|fact|skill","governance":"inspectable","conflict":"llm-arbitrate","description":"First-class memory module inside LangChain — explicit cognitive-science-style typing of memories.","cited_cells":53,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/langmem-langchain--langchain-ai-github-io"},{"id":"langsmith--langchain-com","name":"LangSmith","url":"https://www.langchain.com/langsmith","section":"Memory observability & monitoring","tier":1,"status":"active","last_verified_at":"2026-10-23","type":"Memory mutations as distinct span types","license":"","latest_release":"SDK v0.14.11 (2026-10-16) — platform itself proprietary","deployment":"SaaS cloud (US); Enterprise: dedicated tenant","pricing":"Developer: free (5K traces/mo); Plus: $39/seat/mo (10K traces); Enterprise: custom (SSO/SAML; dedicated support)","mcp_support":"via LangChain MCP adapters","a2a_support":"no data — searched langchain.com/langsmith, docs.smith.langchain.com, docs.smith.langchain.com/observability/concepts; A2A protocol not advertised (OTel-based traces, webhook automations)","backend_storage":"custom (LangChain-managed)","llm_lock":"","github":"","storage":"relational","retrieval":"exact-match","persistence":"long-term","update":"append-only","unit":"episode","governance":"auditable","conflict":"n/a","description":"Memory reads, vector DB retrievals, state changes are distinct span types in traces. RAG eval separates retrieval quality (context precision) from generation quality (faithfulness). Dataset versioning guards against eval drift.","cited_cells":45,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/langsmith--langchain-com"},{"id":"latitude-voyage--latitude-io","name":"Latitude / Voyage","url":"https://latitude.io","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-28","type":"World Engine — separate state-management AI","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free + paid","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"agent-controlled","unit":"episode|scene-graph","governance":"opaque","conflict":"none","description":"AI Dungeon's maker. Voyage (April 2026) launched as a creator platform for AI RPGs. World Engine (5 years in dev) is dedicated AI tracking characters, objects, backstories, relationships independently from the narrative AI; persistent world state evolves with player choices.","cited_cells":32,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/latitude-voyage--latitude-io"},{"id":"lazygraphrag--gh-microsoft-graphrag","name":"LazyGraphRAG","url":"https://www.microsoft.com/en-us/research/blog/lazygraphrag-setting-a-new-standard-for-quality-and-cost/","section":"Retrieval-as-memory hybrids","tier":3,"status":"active","last_verified_at":"2026-10-29","type":"Deferred-indexing GraphRAG variant","license":"MIT","latest_release":"v3.1.2 (2026-08-21)","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"microsoft/graphrag — ~35.1k stars, last commit 2026-08","storage":"graph","retrieval":"graph-traversal|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Performs minimal up-front indexing; defers LLM use until query time. Announced Nov 2024; per a Microsoft Research editor's note (2025-06-06) it ships inside Microsoft Discovery (Azure agentic science platform) and Azure Local services (public preview) rather than in the open-source GraphRAG library, which has no implementation and is largely limited to bug fixes and dependency updates.","cited_cells":26,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/lazygraphrag--gh-microsoft-graphrag"},{"id":"letta-code--gh-letta-ai-letta-code","name":"Letta Code","url":"https://github.com/letta-ai/letta-code","section":"Coding-agent memory","tier":3,"status":"active","last_verified_at":"2026-10-29","type":"Framework","license":"Apache-2.0","latest_release":"last commit 2026-10-22","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/letta-ai/letta-code (3683★)","storage":"file|relational","retrieval":"injection|agentic","persistence":"long-term","update":"agent-controlled|consolidation","unit":"file|skill|episode","governance":"auditable|inspectable","conflict":"agent|document-version","description":"Letta's stateful coding-agent harness: agents rewrite their own memory blocks and skills, version context in git via MemFS (syncable to GitHub) and reflect via a periodic /sleeptime 'dreaming' process. from Letta (MemGPT) team. letta-ai/letta-code is now the main Letta source (harness, terminal UI, App Server, channels); the Letta V1 API server was retired to the archive branch of letta-ai/letta.","cited_cells":15,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/letta-code--gh-letta-ai-letta-code"},{"id":"letta-memgpt--letta-com","name":"Letta / MemGPT","url":"https://www.letta.com/","section":"Dedicated memory layers","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Tiered: core / recall / archival","license":"Apache-2.0","latest_release":"Letta Code v0.34.30 (npm @letta-ai/letta-code, 2026-10-29); PyPI letta v0.34.12 (2026-10-12); last letta-ai/letta server tag v0.16.8 (May 2026)","deployment":"Both","pricing":"Free $0 (3 agents); Pro $20/mo; Developer pay-as-you-go ($0.10/active agent/mo); Teams; Enterprise custom","mcp_support":"native (first-party) — Letta serves & consumes MCP","a2a_support":"not documented publicly","backend_storage":"Postgres + pgvector","llm_lock":"","github":"letta-ai/letta-code — ~3.7k stars, Apache-2.0 (current source: agent harness, terminal UI, App Server); letta-ai/letta (~25.3k stars, not archived) is now a README pointing to letta-code, with the retired Letta V1 API server on its archive branch","storage":"kv|vector","retrieval":"similarity|injection|agentic","persistence":"long-term","update":"agent-controlled","unit":"episode|summary|fact","governance":"inspectable","conflict":"llm-arbitrate","description":"Originated as MemGPT (UC Berkeley), where the model pages its own memory between context, recall and archival tiers. The product now centres on Letta Code, a stateful agent harness whose agents rewrite their own memory, skills and prompts over time.","cited_cells":74,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/letta-memgpt--letta-com"},{"id":"lightrag--arxiv-2410-05779","name":"LightRAG","url":"https://arxiv.org/abs/2410.05779","section":"Retrieval-as-memory hybrids","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Graph-augmented dual-level RAG","license":"MIT","latest_release":"v1.5.7 (2026-09-02)","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"HKUDS/LightRAG — ~39.3k stars, last commit 2026-10","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"accretion","unit":"fact","governance":"plaintext","conflict":"n/a","description":"Hong Kong U graph-RAG variant — dual-level retrieval (local + global) over an entity-relation graph built incrementally from corpus updates.","cited_cells":13,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/lightrag--arxiv-2410-05779"},{"id":"lightrag--gh-hkuds-lightrag","name":"LightRAG","url":"https://github.com/HKUDS/LightRAG","section":"Retrieval-as-memory hybrids","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Dual-level graph retrieval","license":"MIT","latest_release":"v1.5.7 (2026-09-02)","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"HKUDS/LightRAG — ~39.3k stars, last commit 2026-09+","storage":"graph","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Entity-relationship graph + broader-theme retrieval. Incremental update support. Integrates with RAG-Anything for multimodal docs.","cited_cells":28,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/lightrag--gh-hkuds-lightrag"},{"id":"limitless--limitless-ai","name":"Limitless","url":"https://www.limitless.ai/","section":"Voice-first / wearable AI memory","tier":1,"status":"acquired","last_verified_at":"2026-10-29","type":"Wearable pendant + screen capture lifelog","license":"","latest_release":"","deployment":"Managed-only","pricing":"Discontinued — no new sales since 2025-12-05; existing Pendant owners get Unlimited plan free","mcp_support":"via official adapter — Limitless Developer API","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"file|vector","retrieval":"similarity|extraction-pull","persistence":"lifelong","update":"append-only","unit":"episode|document","governance":"user-controllable","conflict":"append","description":"Former Rewind/Limitless lifelog (Pendant wearable + screen/audio capture with timeline search). Acquired by Meta (2025-12); Pendant no longer sold as of 2025-12-05, service withdrawn from the EU, UK, Brazil, China, Israel, South Korea and Turkey on 2025-12-05, desktop/web recording ended and Rewind capture disabled from 2025-12-19; existing Pendant customers supported throughout 2026.","cited_cells":38,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/limitless--limitless-ai"},{"id":"lindy-ai-memory--docs-lindy-ai","name":"Lindy AI Memory","url":"https://docs.lindy.ai/teammate/memory","section":"Framework-embedded memory","tier":1,"status":"deprecated","last_verified_at":"2026-10-23","type":"Selective KV memory injected into prompt","license":"","latest_release":"","deployment":"Managed cloud only (SaaS)","pricing":"Free: $50 in credits for 7 days, no card; Team $29.99/user/mo (3,000 credits, extra $10 per 1,000); Enterprise custom","mcp_support":"yes — MCP client: connects to hosted streamable-HTTP MCP servers and builds actions from their tools","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"multi-vendor","github":"","storage":"kv","retrieval":"injection","persistence":"cross-session","update":"extraction","unit":"fact","governance":"user-controllable","conflict":"llm-arbitrate","description":"Memory is a file explorer for what Lindy knows: files it creates or uses, files users upload, skills (.skills/ folders with SKILL.md) and durable knowledge in a .memory/ area. Scoped Personal (owner only), Team (admins edit, members read-only) and System (read-only); every save is versioned and restorable.","cited_cells":44,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/lindy-ai-memory--docs-lindy-ai"},{"id":"linkedin-cognitive-memory-agent--linkedin-com","name":"LinkedIn Cognitive Memory Agent","url":"https://www.linkedin.com/blog/engineering/ai/the-linkedin-generative-ai-application-tech-stack-personalization-with-cognitive-memory-agent","section":"Platform-provider memory","tier":1,"status":"active","last_verified_at":"2026-10-28","type":"Episodic + semantic + procedural (shared)","license":"","latest_release":"","deployment":"Managed-only (internal LinkedIn production infrastructure)","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|graph","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction","unit":"episode|fact|skill","governance":"opaque","conflict":"none","description":"Production memory infrastructure powering LinkedIn's Hiring Assistant and other GenAI apps. Three-layer (episodic/semantic/procedural) shared memory substrate across multi-agent systems; recent-context retrieval + semantic search + summarisation-based compaction. April 2026.","cited_cells":41,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/linkedin-cognitive-memory-agent--linkedin-com"},{"id":"liveperson-conversation-context-service--developers-liveperson-com","name":"LivePerson Conversation Context Service","url":"https://developers.liveperson.com/conversation-orchestrator-conversation-context-service-overview.html","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Named cloud-based shared session-state KV","license":"","latest_release":"","deployment":"searched not found","pricing":"searched not found","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"kv","retrieval":"exact-match","persistence":"cross-session","update":"overwrite","unit":"fact","governance":"inspectable","conflict":"last-write-wins","description":"Architecturally notable — a named, documented service. Shared key-value store any agent (bot or human) can read/write across a customer's conversational journey, reducing repetition and enabling continuity.","cited_cells":25,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/liveperson-conversation-context-service--developers-liveperson-com"},{"id":"llamaindex-memory--llamaindex-ai","name":"LlamaIndex Memory","url":"https://www.llamaindex.ai/","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"FIFO short-term + Memory Blocks","license":"MIT","latest_release":"v0.14.25 (2026-09-21)","deployment":"Both","pricing":"Free + paid","mcp_support":"via official adapter — llama-index-tools-mcp","a2a_support":"supported — LlamaIndex A2A agent server","backend_storage":"pluggable","llm_lock":"any (BYO)","github":"run-llama/llama_index — ~52.6k stars, last commit 2026-09 or later","storage":"kv","retrieval":"injection","persistence":"cross-session","update":"append-only|consolidation","unit":"episode|summary","governance":"inspectable","conflict":"append-only","description":"Two-tier model: ChatMessage FIFO queue with overflow flushed into long-term Memory Block objects. Memory class replacing older ChatMemoryBuffer.","cited_cells":60,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/llamaindex-memory--llamaindex-ai"},{"id":"local-memory-mcp--gh-studiomeyer-io-local-memory-mcp","name":"local-memory-mcp","url":"https://github.com/studiomeyer-io/local-memory-mcp","section":"Claude Code memory mechanisms","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"SQLite + FTS5 with decision-tracking","license":"MIT","latest_release":"v2.4.3 (2026-08-19)","deployment":"local — SQLite + FTS5; zero cloud dependency; no API keys required","pricing":"free / open source (MIT); no paid tier","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"studiomeyer-io/local-memory-mcp — ~13 stars, 86 commits","storage":"relational","retrieval":"exact-match|similarity","persistence":"long-term","update":"overwrite|append-only","unit":"fact|document","governance":"inspectable","conflict":"append","description":"SQLite + FTS5 + BM25 ranking. Stores learnings, decisions with reasoning and alternatives , entities, observations. Loads context from last 3 sessions. Only server that captures why a decision was made + alternatives considered.","cited_cells":42,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/local-memory-mcp--gh-studiomeyer-io-local-memory-mcp"},{"id":"logseq--logseq-com","name":"Logseq","url":"https://logseq.com/","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"active","last_verified_at":"2026-10-29","type":"Local Markdown outliner + AI plugins","license":"AGPL-3.0","latest_release":"2.0.2 (2026-10-07) — desktop app, labelled beta testing","deployment":"Both (desktop app local-first; optional sync add-on planned)","pricing":"Free (open-source; all local features free); Logseq Sync beta available to Open Collective backers/sponsors on a $5 or $15/month recurring donation, normal subscription promised at public launch; DB-version RTC sync in alpha","mcp_support":"","a2a_support":"","backend_storage":"file (Markdown/Org-mode)","llm_lock":"","github":"logseq/logseq — ~45k stars, last commit not confirmed","storage":"file|graph","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"append-only","unit":"document|fact","governance":"user-controllable","conflict":"append","description":"Open-source local-first outliner. Notes stored as Markdown / Org-mode on disk. AI integration via community plugins (e.g. logseq-ai for chatting with the graph). Graph view + backlinks; daily-journal-driven workflow.","cited_cells":44,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/logseq--logseq-com"},{"id":"longrag--arxiv-2406-15319","name":"LongRAG","url":"https://arxiv.org/abs/2406.15319","section":"Retrieval-as-memory hybrids","tier":4,"status":"research-code","last_verified_at":"2026-10-22","type":"Long retrieval-unit grouping","license":"MIT","latest_release":"no-release no-release","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"248★ Python","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"read-only","unit":"document","governance":"inspectable","conflict":"n/a","description":"Groups related Wikipedia documents into 4K-token units. Reduces NQ corpus 22M→600K units, HotpotQA 5M→500K. Hands longer passages to a long-context reader (GPT-4o).","cited_cells":27,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/longrag--arxiv-2406-15319"},{"id":"lovable--lovable-dev","name":"Lovable","url":"https://lovable.dev","section":"Coding-agent memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"/docs folder + Knowledge Panel","license":"","latest_release":"","deployment":"Managed (editor/agent are Lovable-hosted); Enterprise private preview runs generated apps + data plane in customer AWS/Azure account","pricing":"Free + paid","mcp_support":"yes — chat connectors incl. custom MCP servers on all plans; first-party Lovable MCP server at mcp.lovable.dev (OAuth) for external AI clients","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"none","description":"Long-term context stored as files in /docs that the agent reads on each session. Agent Mode (default since July 2025) holds working context within session but doesn't persist cross-session beyond filesystem state.","cited_cells":47,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/lovable--lovable-dev"},{"id":"luminance-institutional-memory--luminance-com","name":"Luminance Institutional Memory","url":"https://www.luminance.com/press-releases/luminance-launches-new-legal-ai-with-institutional-memory-addressing-enterprise-amnesia-and-giving-legal-teams-30-of-their-time-back","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Persistent negotiation history + accepted-risk registry","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|relational","retrieval":"similarity|injection","persistence":"long-term","update":"extraction|overwrite","unit":"episode|document|fact","governance":"auditable","conflict":"overwrite","description":"Jan 2026 platform relaunch. Persists negotiation history, prior decisions, accepted risk positions across all contracts; new terms align with what the org has already agreed. Branded as the antidote to \"enterprise amnesia.\"","cited_cells":36,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/luminance-institutional-memory--luminance-com"},{"id":"m-flow-m-flow--gh-flowelement-xinliuyuansu-m-flow","name":"m_flow (M-Flow)","url":"https://github.com/FlowElement-xinliuyuansu/m_flow","section":"Retrieval-as-memory hybrids","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"Apache-2.0","latest_release":"last commit 2026-08-03","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/FlowElement-xinliuyuansu/m_flow (4512★)","storage":"hybrid|graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"extraction","unit":"episode|fact|skill","governance":"inspectable","conflict":"n/a","description":"Cognitive memory engine modelled on biology that presents itself as a new paradigm for graph RAG for agents. #13 on GitHub agent-memory topic","cited_cells":15,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/m-flow-m-flow--gh-flowelement-xinliuyuansu-m-flow"},{"id":"m3-agent-bytedance--gh-bytedance-seed-m3-agent","name":"M3-Agent (ByteDance)","url":"https://github.com/ByteDance-Seed/m3-agent","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Episodic + semantic multimodal memory graph","license":"Apache-2.0","latest_release":"no releases","deployment":"Self-hosted (research code; API config + memory graph files)","pricing":"not applicable — academic research code (no commercial pricing)","mcp_support":"not documented publicly","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"ByteDance-Seed/m3-agent — ~1.4k stars, last commit not confirmed","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"extraction","unit":"episode|fact","governance":"opaque","conflict":"n/a","description":"ByteDance Seed + Zhejiang Univ + SJTU. Multimodal agent with long-term memory as entity-centric graph built from live video + audio streams. Parallel memorisation + control processes. Ships M3-Bench evaluation dataset.","cited_cells":36,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/m3-agent-bytedance--gh-bytedance-seed-m3-agent"},{"id":"magicschool--magicschool-ai","name":"MagicSchool","url":"https://www.magicschool.ai/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Teacher tooling — no student-facing memory by design","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free (teachers); Plus $12.99/mo or $99.96/yr; Enterprise (districts) custom","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"file","retrieval":"injection","persistence":"session","update":"read-only","unit":"document","governance":"opaque","conflict":"n/a","description":"Teacher productivity platform (80+ teacher tools, 50+ student tools) for planning, differentiation, IEP generation, feedback. Does not maintain a cross-session learner model. Teachers input student context manually each session; district deployments can pre-populate fields via RAG over uploaded curriculum/policy documents.","cited_cells":32,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/magicschool--magicschool-ai"},{"id":"mastra-memory--mastra-ai","name":"Mastra Memory","url":"https://mastra.ai/docs/memory/overview","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"LibSQL + semantic recall","license":"Apache-2.0 (core; ee/ dirs under source-available Mastra Enterprise License)","latest_release":"@mastra/core v1.79.0 (2026-10-21)","deployment":"Both","pricing":"Free + paid","mcp_support":"native (first-party) — Mastra MCP server + client","a2a_support":"supported — Mastra A2A protocol","backend_storage":"pluggable (Postgres, LibSQL, etc.)","llm_lock":"any (BYO; OpenAI/Anthropic/Google/local)","github":"mastra-ai/mastra — ~27.1k stars, last commit 2026-10","storage":"relational|vector","retrieval":"similarity|injection","persistence":"cross-session","update":"append-only","unit":"episode","governance":"inspectable","conflict":"none","description":"TypeScript-first agent framework. Default memory: LibSQL store + on-device fastembed-js for semantic recall.","cited_cells":65,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mastra-memory--mastra-ai"},{"id":"mcp-knowledge-graph--gh-shaneholloman-mcp-knowledge-graph","name":"mcp-knowledge-graph","url":"https://github.com/shaneholloman/mcp-knowledge-graph","section":"Claude Code memory mechanisms","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Stable local fork of official MCP memory","license":"MIT","latest_release":"v1.3.2 2025-12-22","deployment":"local (Node.js MCP server; local JSON file storage)","pricing":"free / open source (MIT)","mcp_support":"native (first-party) — fork of official memory server","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"shaneholloman/mcp-knowledge-graph — ~889 stars, last commit n/a","storage":"graph","retrieval":"graph-traversal","persistence":"long-term","update":"overwrite|append-only","unit":"fact","governance":"inspectable","conflict":"llm-arbitrate","description":"KG-based local fork of @modelcontextprotocol/server-memory . Avoids ephemeral npx installs; supports named databases for topic scoping; user-controlled memory path. Drop-in for teams that want the official KG model without npx churn.","cited_cells":41,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mcp-knowledge-graph--gh-shaneholloman-mcp-knowledge-graph"},{"id":"mcp-memory-puliczek--gh-puliczek-mcp-memory","name":"mcp-memory (Puliczek)","url":"https://github.com/Puliczek/mcp-memory","section":"Claude Code memory mechanisms","tier":2,"status":"stale","last_verified_at":"2026-10-28","type":"Cloudflare-native vector memory","license":"Unlicensed","latest_release":"no releases","deployment":"serverless — Cloudflare Workers + Vectorize + D1; one-click Cloudflare deploy","pricing":"free — Cloudflare free tier: 1k memories / ~28k queries per month; no paid tier","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"Puliczek/mcp-memory — ~168 stars, last commit 2025-04","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"overwrite|append-only","unit":"chunk","governance":"inspectable","conflict":"llm-arbitrate","description":"Embeddings via @cf/baai/bge-m3 ; Vectorize + D1; per-user namespace isolation. Runs entirely on Cloudflare Workers — no separate server.","cited_cells":41,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mcp-memory-puliczek--gh-puliczek-mcp-memory"},{"id":"mcp-memory-service-doobidoo--gh-doobidoo-mcp-memory-service","name":"mcp-memory-service (doobidoo)","url":"https://github.com/doobidoo/mcp-memory-service","section":"Claude Code memory mechanisms","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"KG + vector hybrid + autonomous consolidation","license":"Apache-2.0","latest_release":"v11.15.0 (2026-10-03)","deployment":"Local/self-hosted — SQLite-vec local; Cloudflare, Hybrid or Milvus backends; no managed cloud","pricing":"free / open source (Apache-2.0); no paid tier","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"doobidoo/mcp-memory-service — ~2.0k stars, last release 2026-10","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"extraction|consolidation","unit":"fact|episode","governance":"inspectable","conflict":"llm-arbitrate","description":"SQLite-Vec default, Cloudflare hybrid for production. Typed edges (causes, fixes, contradicts). Autonomous consolidation compresses old memories. SSE real-time notifications + D3.js visualisation dashboard. Agent-pipeline focused (LangGraph, CrewAI, AutoGen).","cited_cells":45,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mcp-memory-service-doobidoo--gh-doobidoo-mcp-memory-service"},{"id":"mem-ai--mem-ai","name":"Mem.ai","url":"https://mem.ai/","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"AI-first note-taking","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free; Plus $9/mo; Pro $29/mo; Pro 2×/5×/10×/20× at $49/$99/$149/$199/mo","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"none","description":"Note-taking with no folders; AI-chat over notes. Focused on connections between notes rather than hierarchy.","cited_cells":40,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mem-ai--mem-ai"},{"id":"mem0--mem0-ai","name":"Mem0","url":"https://mem0.ai/","section":"Dedicated memory layers","tier":1,"status":"active","last_verified_at":"2026-10-23","type":"Vector + graph + KV (hybrid)","license":"Apache-2.0","latest_release":"v2.2.1 (2026-09-25) — mem0ai on PyPI","deployment":"Both","pricing":"Free + paid","mcp_support":"yes — official hosted Mem0 MCP server (docs.mem0.ai/platform/mem0-mcp); the self-hosted mem0-mcp repo was archived 2026-03-24","a2a_support":"not documented publicly","backend_storage":"hybrid (vector + graph + KV)","llm_lock":"","github":"mem0ai/mem0 — ~67.3k stars, last commit 2026-09+","storage":"vector|graph|kv","retrieval":"similarity|graph-traversal|extraction-pull","persistence":"long-term","update":"extraction","unit":"fact|episode","governance":"opaque","conflict":"llm-arbitrate","description":"Universal memory layer for AI agents. Three concurrent stores (vector + graph + KV); LLM-extracted facts; concurrent retrieval via ThreadPoolExecutor.","cited_cells":81,"third_party_cells":8,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mem0--mem0-ai"},{"id":"mem0-mcp-official--gh-mem0ai-mem0-mcp","name":"Mem0 MCP (official)","url":"https://github.com/mem0ai/mem0-mcp","section":"Claude Code memory mechanisms","tier":1,"status":"archived","last_verified_at":"2026-10-29","type":"Hybrid vector + graph via cloud API","license":"Apache-2.0","latest_release":"no releases","deployment":"superseded — use Mem0's hosted MCP endpoint (mcp.mem0.ai/mcp) or the mem0-mcp-server PyPI package","pricing":"free tier (Hobby: 10K memories 1K retrieval/mo); Starter $19/mo; Pro $249/mo (graph memory included); Enterprise custom","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"hybrid","llm_lock":"","github":"mem0ai/mem0-mcp — ~658 stars, archived 2026-03-24","storage":"vector|graph","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction","unit":"fact|episode","governance":"opaque","conflict":"llm-arbitrate","description":"Archived 2026-03-24 (read-only); README now points to Mem0's official cloud-hosted MCP server at mcp.mem0.ai/mcp. Mem0's managed cloud API exposed as MCP tools. Semantic search + CRUD on memories scoped by user / agent / session. Cloud-hosted MCP endpoint — no local infrastructure required.","cited_cells":51,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mem0-mcp-official--gh-mem0ai-mem0-mcp"},{"id":"mem0-security-openmemory--mem0-ai","name":"Mem0 Security / OpenMemory","url":"https://trust.mem0.ai/","section":"Memory governance, privacy & safety","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"SOC 2 + HIPAA + zero-trust + BYOK","license":"Apache-2.0","latest_release":"v2.2.1 (2026-09-25)","deployment":"SaaS cloud; Enterprise VPC; OpenMemory hosted or self-hosted (Docker)","pricing":"Hobby free (10K adds/1K retrievals/mo); Starter $19/mo; Pro $249/mo; Enterprise custom","mcp_support":"native (first-party)","a2a_support":"no data — searched mem0.ai/security, docs.mem0.ai, docs.mem0.ai/platform/overview; A2A protocol not advertised","backend_storage":"hybrid","llm_lock":"","github":"mem0ai/mem0 — ~67.3k stars, last release 2026-09","storage":"vector|graph","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction","unit":"fact","governance":"auditable","conflict":"llm-arbitrate","description":"Commercial Mem0 ships SOC 2 / HIPAA, zero-trust access controls, BYOK encryption, real-time monitoring, audit logs and workspace governance as defaults (Trust Center at trust.mem0.ai). For privacy-first deployments the self-hosted path is now the Mem0 open-source server (REST API + dashboard via Docker Compose with pgvector Postgres, admin setup wizard, per-user API keys, request audit log). As of Oct 2026 OpenMemory itself is a hosted MCP service and its old local Docker variant is no longer in the repo.","cited_cells":66,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mem0-security-openmemory--mem0-ai"},{"id":"memanto-moorcheh--memanto-ai","name":"Memanto (Moorcheh)","url":"https://memanto.ai/","section":"Dedicated memory layers","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"MIT","latest_release":"last commit 2026-10-21","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/moorcheh-ai/memanto (2337★)","storage":"proprietary|vector","retrieval":"similarity|injection","persistence":"long-term","update":"consolidation|extraction|agent-controlled","unit":"fact","governance":"auditable|editable|deterministic","conflict":"human-arbitrate|bi-temporal","description":"A 'memory agent' that manages other agents' memories, deciding what to keep, what conflicts, what expires and who needs to know; runs locally (Docker/Ollama) or as a free cloud. self-reported 89.8% LongMemEval / 87.1% LoCoMo","cited_cells":16,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memanto-moorcheh--memanto-ai"},{"id":"memary--gh-kingjulio8238-memary","name":"Memary","url":"https://github.com/kingjulio8238/Memary","section":"Dedicated memory layers","tier":2,"status":"inactive","last_verified_at":"2026-10-22","type":"Long-term memory for autonomous agents","license":"MIT","latest_release":"v0.1.5 2024-10-22","deployment":"Self-hosted (OSS; no cloud service found)","pricing":"OSS (MIT); no commercial pricing","mcp_support":"not documented publicly","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"kingjulio8238/Memary — ~2.6k stars, last commit n/a","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"extraction","unit":"episode|fact","governance":"user-controllable","conflict":"none","description":"Open-source long-term memory framework for autonomous agents.","cited_cells":48,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memary--gh-kingjulio8238-memary"},{"id":"memmachine--gh-memmachine-memmachine","name":"MemMachine","url":"https://github.com/MemMachine/MemMachine","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Multi-layered memory system","license":"Apache-2.0","latest_release":"v0.3.9 (2026-05-18)","deployment":"Both — self-hosted, or managed MemMachine Platform","pricing":"OSS free (Apache-2.0); managed MemMachine Platform with free account (paid tiers not listed in README)","mcp_support":"native (first-party) — MemMachine MCP","a2a_support":"not documented publicly","backend_storage":"hybrid","llm_lock":"","github":"MemMachine/MemMachine — ~3.2k stars","storage":"vector|graph","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction","unit":"episode|fact","governance":"inspectable","conflict":"none","description":"Multi-layered open-source memory system for AI agents.","cited_cells":47,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memmachine--gh-memmachine-memmachine"},{"id":"memobase--gh-memodb-io-memobase","name":"Memobase","url":"https://github.com/memodb-io/memobase","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-29","type":"User profile + event timeline","license":"Apache-2.0","latest_release":"memobase Python SDK 0.0.27 (PyPI, 2025-11-08); no tagged GitHub releases","deployment":"Both (self-hosted Docker + Memobase Cloud managed)","pricing":"OSS self-hosted free (Apache-2.0); Memobase Cloud free and pay-as-you-go tiers advertised in the README, but memobase.io (incl. pricing and login) returns 404 and app.memobase.io is down as of 2026-10-29","mcp_support":"native (first-party) — memobase MCP","a2a_support":"not supported","backend_storage":"Postgres + pgvector","llm_lock":"","github":"memodb-io/memobase — ~2.8k stars, last commit ~2026-01","storage":"relational|vector","retrieval":"injection|similarity","persistence":"long-term","update":"extraction|append-only","unit":"profile|episode","governance":"inspectable","conflict":"none","description":"User-profile-based long-term memory. Per-user profile + event timeline. API + Python/Node/Go SDKs + MCP server. Profile-centric rather than fact-centric.","cited_cells":57,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memobase--gh-memodb-io-memobase"},{"id":"memorag--arxiv-2409-05591","name":"MemoRAG","url":"https://arxiv.org/abs/2409.05591","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"Memory-augmented RAG","license":"Apache-2.0","latest_release":"n/a","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"qhjqhj00/MemoRAG — ~2.3k stars, 68 commits","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"accretion","unit":"fact","governance":"plaintext","conflict":"n/a","description":"RAG variant with a long-context global memory model that drafts answer clues used to guide passage retrieval — improves recall in noisy multi-hop QA.","cited_cells":11,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memorag--arxiv-2409-05591"},{"id":"memorax-ai--eu-36kr-com","name":"MemoraX AI","url":"https://www.memoraxai.com/","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-29","type":"Endogenous / in-model memory (no external bank)","license":"MIT (MemoraX Code client); hosted memory platform proprietary","latest_release":"MemoraX Code 0.1.21 (npm @memorax/memorax-code, 2026-10-21)","deployment":"searched not found","pricing":"searched not found","mcp_support":"not documented publicly","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"memorax-ai/memorax-code (MemoraX Code, MIT); org github.com/memorax-ai also hosts ScriptMem","storage":"parametric","retrieval":"parametric-recall","persistence":"parametric-permanent","update":"parametric-edit","unit":"weight","governance":"opaque","conflict":"none","description":"Shenzhen. Argues attaching external vector stores is architecturally flawed; trains memory as an intrinsic capability via Agentic RL. Plans B2B enterprise KM and C2C personalisation in 12 months. Founded by Tianjin University AI professor. Company site memoraxai.com (operated by 深圳忆纪元科技有限公司) describes textual, work/coding and multimodal memory infrastructure; ships MemoraX Code, an MIT-licensed cross-agent coding memory for Codex and Claude Code (npm @memorax/memorax-code, first published 2026-08), plus public evals LoCoMo-Refined and ScriptMem.","cited_cells":24,"third_party_cells":17,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memorax-ai--eu-36kr-com"},{"id":"memori--gh-memorilabs-memori","name":"Memori","url":"https://github.com/MemoriLabs/Memori","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Structured persistent memory layer","license":"Apache-2.0","latest_release":"v3.3.6 (2026-05-28)","deployment":"Both (Memori Cloud zero-config + self-hosted OSS)","pricing":"Free + paid (Memori Cloud; tiered plans for builders / growing / production teams)","mcp_support":"yes — first-party hosted remote MCP endpoint (api.memorilabs.ai/mcp) for Claude Code, Cursor, Codex etc.","a2a_support":"not documented publicly","backend_storage":"SQL (SQLite/Postgres/MySQL)","llm_lock":"","github":"MemoriLabs/Memori — ~17.1k stars, last commit n/a","storage":"relational|vector","retrieval":"similarity|injection","persistence":"cross-session","update":"extraction","unit":"fact|episode","governance":"inspectable","conflict":"append","description":"Structured, persistent memory layer for AI agents. Open source.","cited_cells":49,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memori--gh-memorilabs-memori"},{"id":"memori-gibsonai--gh-gibsonai-memori","name":"Memori (GibsonAI) (now Memori Labs)","url":"https://github.com/MemoriLabs/Memori","section":"Dedicated memory layers","tier":3,"status":"renamed","last_verified_at":"2026-10-23","type":"OSS memory layer (SQL-first)","license":"Apache-2.0","latest_release":"v3.3.6 (2026-05-28)","deployment":"Both — OSS SDK self-hosted; Memori Cloud managed; enterprise single-tenant/VPC/on-prem","pricing":"OSS free (Apache-2.0); Memori Cloud API key with free developer tier (rate-limited); enterprise by contact","mcp_support":"yes — hosted remote MCP endpoint for Claude Code, Cursor, Codex, Warp, Antigravity (no SDK needed)","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"MemoriLabs/Memori — ~17.1k stars, last commit 2026-05","storage":"kv","retrieval":"similarity|extraction-pull","persistence":"long-term","update":"extraction","unit":"fact","governance":"plaintext","conflict":"n/a","description":"Now Memori by Memori Labs (repo moved to MemoriLabs/Memori; adds managed Memori Cloud and BYODB). SQL-native memory engine for AI agents — single library, opt-in conscious/auto modes, structured Pydantic outputs. Aims to be 'one line of code' to add memory to any LLM workflow.","cited_cells":20,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memori-gibsonai--gh-gibsonai-memori"},{"id":"memoria-matrixorigin--medium-com","name":"Memoria (MatrixOrigin)","url":"https://medium.com/@matrixorigin-database/benchmarking-memoria-on-longmemeval-strong-memory-retrieval-clear-reader-separation-ee6c89c75d76","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Retrieval-strong memory layer","license":"Apache-2.0","latest_release":"v0.5.2 (Sep 11; page shows no year, presumably 2026)","deployment":"Both — Memoria Cloud (thememoria.ai, remote mode) or self-hosted Docker Compose with MatrixOne (embedded mode); Kubernetes documented","pricing":"searched not found","mcp_support":"yes — native MCP server (memoria mcp) for any MCP-compatible agent","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"matrixorigin/memoria — ~607 stars","storage":"hybrid","retrieval":"similarity","persistence":"long-term","update":"extraction","unit":"fact","governance":"inspectable","conflict":"none","description":"Memory layer from MatrixOrigin. Strong-retrieval / clear-reader-separation design.","cited_cells":34,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memoria-matrixorigin--medium-com"},{"id":"memories-ai-lvmm-2-0-project-luci--memories-ai","name":"Memories.ai LVMM 2.0 / Project LUCI","url":"https://memories.ai/blog/collaboration-with-qualcomm","section":"Voice-first / wearable AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"On-device Large Visual Memory Model for wearables","license":"","latest_release":"","deployment":"Hybrid (on-device LVMM 2.0 + cloud companion)","pricing":"Free 100 credits/mo; Plus $15/mo (5,000 credits, annual); Enterprise custom (incl. on-prem); usage API e.g. storage $0.03/GB-mo, search $0.10/1K queries","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"Qualcomm SoC + on-device storage + cloud sync","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"append-only","unit":"episode","governance":"user-controllable","conflict":"none","description":"Developer-first reference platform for AI wearables. LVMM 2.0 captures visual data through video, encodes it into frames on-device, and stores moments as searchable memories; companion app surfaces past conversations/decisions/experiences. Hardware-level security via Qualcomm partnership; cloud sync for enterprise tiers.","cited_cells":37,"third_party_cells":11,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memories-ai-lvmm-2-0-project-luci--memories-ai"},{"id":"memoripy--gh-caspianmoon-memoripy","name":"Memoripy","url":"https://github.com/caspianmoon/memoripy","section":"Dedicated memory layers","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"Python OSS memory library (short+long term)","license":"Apache-2.0","latest_release":"v0.1.2 (2024-11-28) — PyPI","deployment":"searched not found","pricing":"free / open source (Apache-2.0)","mcp_support":"yes — official stdio MCP server via [mcp] extra (memoripy mcp)","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"caspianmoon/memoripy — ~695 stars, 32 commits on master","storage":"vector","retrieval":"similarity","persistence":"long-term|session","update":"accretion","unit":"fact|episode","governance":"plaintext","conflict":"n/a","description":"Memory layer for AI assistants — semantic short-term and long-term memory with concept-graph linking and decay/reinforcement.","cited_cells":15,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memoripy--gh-caspianmoon-memoripy"},{"id":"memory-lancedb-pro-openclaw-plugin--gh-cortexreach-memory-lancedb-pro","name":"memory-lancedb-pro (OpenClaw plugin)","url":"https://github.com/CortexReach/memory-lancedb-pro","section":"Framework-embedded memory","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"MIT","latest_release":"last commit 2026-10-21","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/CortexReach/memory-lancedb-pro (4452★)","storage":"column|vector|file","retrieval":"hybrid-rerank","persistence":"long-term","update":"extraction|consolidation","unit":"fact|profile|summary","governance":"editable|user-controllable","conflict":"llm-arbitrate","description":"Enhanced LanceDB memory plugin for OpenClaw that adds hybrid vector + BM25 retrieval, cross-encoder reranking, multi-scope isolation and a management CLI. #21 in TeleAI Awesome-Agent-Memory","cited_cells":15,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memory-lancedb-pro-openclaw-plugin--gh-cortexreach-memory-lancedb-pro"},{"id":"memory-store--ycombinator-com","name":"Memory Store (now MemStore)","url":"https://mem.store/","section":"Dedicated memory layers","tier":2,"status":"renamed","last_verified_at":"2026-10-23","type":"Cross-agent shared episodic + preference memory","license":"","latest_release":"","deployment":"Managed-only (hosted; no self-host option documented)","pricing":"$150 per team member/mo after 7-day trial (single plan)","mcp_support":"yes — is an MCP server (checkin/record/recall tools) at mem.store/mcp for Claude, Claude Code, ChatGPT, Codex, any MCP client","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"cross-session","update":"extraction","unit":"episode|fact","governance":"opaque","conflict":"none","description":"Renamed MemStore (memory.store now redirects to mem.store); YC Spring 2026. Shared memory layer delivered as an MCP server: a team's AI agents record decisions/progress across sessions and recall from teammates' stores with sources attached; daily email rollup.","cited_cells":24,"third_party_cells":14,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memory-store--ycombinator-com"},{"id":"memorybear--memorybear-ai","name":"MemoryBear","url":"https://www.memorybear.ai/","section":"Dedicated memory layers","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"Apache-2.0","latest_release":"last commit 2026-10-14","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/SuanmoSuanyangTechnology/MemoryBear (8260★)","storage":"graph|relational|vector","retrieval":"hybrid-rerank|graph-traversal","persistence":"long-term","update":"extraction|consolidation|evict-oldest","unit":"fact","governance":"editable|inspectable","conflict":"human-arbitrate","description":"Open-source memory system modelled on human cognition, with perception, extraction, association and forgetting stages; also offered as a hosted product. #18 in TeleAI Awesome-Agent-Memory","cited_cells":15,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memorybear--memorybear-ai"},{"id":"memoryvla--arxiv-2508-19236","name":"MemoryVLA","url":"https://arxiv.org/abs/2508.19236","section":"Vertical / domain-specific AI memory","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Perceptual + Cognitive Memory Bank (ICLR 2026)","license":"Unlicensed","latest_release":"no releases","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"shihao1895/MemoryVLA — ~361 stars, last commit n/a (README news 2026-06)","storage":"hybrid","retrieval":"similarity|injection","persistence":"long-term","update":"consolidation","unit":"episode","governance":"inspectable","conflict":"append","description":"Two-tier memory bank on a VLA backbone. Perceptual MB holds low-level visual details; Cognitive MB holds high-level semantic summaries; working-memory module retrieves and fuses both into observation tokens before action prediction. Hippocampal + semantic-gist inspiration.","cited_cells":32,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memoryvla--arxiv-2508-19236"},{"id":"memsearch-zilliz--gh-zilliztech-memsearch","name":"memsearch (Zilliz)","url":"https://github.com/zilliztech/memsearch","section":"Coding-agent memory","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"MIT","latest_release":"last commit 2026-10-22","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/zilliztech/memsearch (2775★)","storage":"file|vector","retrieval":"hybrid-rerank","persistence":"long-term","update":"append-only|extraction|consolidation","unit":"chunk|summary","governance":"editable|plaintext","conflict":"append-only","description":"Zilliz's persistent memory layer shared across agents like Claude Code and Codex: Markdown files are the source of truth, and Milvus indexes them for search. from the Milvus vendor","cited_cells":15,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memsearch-zilliz--gh-zilliztech-memsearch"},{"id":"memu-nevamind--gh-nevamind-ai-memu","name":"memU (NevaMind)","url":"https://github.com/NevaMind-AI/memU","section":"Dedicated memory layers","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"Apache-2.0","latest_release":"last commit 2026-10-16","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/NevaMind-AI/memU (14556★)","storage":"file|relational|vector","retrieval":"similarity|agentic","persistence":"long-term","update":"agent-controlled|extraction","unit":"document|skill","governance":"editable","conflict":"agent","description":"Lightweight agent memory system that builds a shared, file-like 'memory wiki' across sessions, agents and devices and can extract reusable skills from agent history; also offered as a hosted service (memu.pro). #9 on GitHub agent-memory topic","cited_cells":16,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memu-nevamind--gh-nevamind-ai-memu"},{"id":"memvid--gh-memvid-memvid","name":"Memvid","url":"https://github.com/memvid/memvid","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Single-file append-only Smart Frames","license":"Apache-2.0","latest_release":"v2.0.140 (2026-05-27)","deployment":"Self-hosted only (single .mv2 file; no cloud service)","pricing":"OSS (Apache-2.0); free — no commercial pricing found","mcp_support":"native (first-party) — claude-brain MCP","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"memvid/memvid — ~16.6k stars, last commit 2026-05","storage":"file","retrieval":"similarity","persistence":"long-term","update":"append-only","unit":"chunk|document","governance":"inspectable","conflict":"append","description":"Single-file memory layer (one .mv2 file). No DB, no server. Append-only sequence of immutable Smart Frames with timestamps + checksums. Native Rust core (rewritten from Python).","cited_cells":51,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memvid--gh-memvid-memvid"},{"id":"memwal-walrus--decrypt-co","name":"MemWal (now Walrus Memory)","url":"https://decrypt.co/365834/agentic-memory-walrus-takes-on-ais-next-big-bottleneck","section":"Dedicated memory layers","tier":2,"status":"renamed","last_verified_at":"2026-10-29","type":"Decentralised / verifiable on-chain agent memory","license":"Apache-2.0","latest_release":"@mysten-incubation/memwal 0.1.5 (Aug 26; page shows no year, presumably 2026)","deployment":"Decentralised blockchain storage (Sui blockchain via Walrus protocol; SDK in beta)","pricing":"Comparable to centralized models (per Mysten Labs; $WAL token-based usage)","mcp_support":"not documented publicly","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"MystenLabs/MemWal — ~117 stars, beta","storage":"proprietary","retrieval":"exact-match","persistence":"lifelong","update":"append-only","unit":"document|episode","governance":"auditable","conflict":"append","description":"Walrus is a Sui-blockchain decentralised storage protocol; MemWal is its AI-facing SDK that stores encrypted agent memory on-chain — portable, verifiable, shareable across agent systems with no single-vendor lock-in. Targets production agents needing auditability. Renamed Walrus Memory (memwal.ai redirects to memory.walrus.xyz); still in beta, SDK published as @mysten-incubation/memwal.","cited_cells":40,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memwal-walrus--decrypt-co"},{"id":"meta-ai-memory--meta-com","name":"Meta AI Memory","url":"https://about.fb.com/news/2025/01/building-toward-a-smarter-more-personalized-assistant/","section":"Platform-provider memory","tier":1,"status":"active","last_verified_at":"2026-10-23","type":"Cross-platform via Accounts Center","license":"","latest_release":"","deployment":"Managed-only","pricing":"Meta AI free; Meta One Core $7.99/mo and Premium $19.99/mo add more AI generation capacity (memory not tier-gated per announcement)","mcp_support":"searched not found","a2a_support":"not supported","backend_storage":"custom (Meta-managed)","llm_lock":"","github":"","storage":"kv","retrieval":"injection","persistence":"long-term","update":"extraction","unit":"fact","governance":"user-controllable","conflict":"llm-arbitrate","description":"Meta AI remembers details users share in 1:1 chats (or asks it to remember) and uses them to personalise later replies; memory launched in Facebook, Messenger and WhatsApp in the US and Canada, with separate personalised responses drawing on Facebook and Instagram account info. Group chats are excluded.","cited_cells":58,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/meta-ai-memory--meta-com"},{"id":"meta-memory--arxiv-2509-20754","name":"Meta-Memory","url":"https://arxiv.org/abs/2509.20754","section":"Vertical / domain-specific AI memory","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"Semantic-spatial dual-modality memory + multi-hop retrieval","license":"MIT","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"ItsBaymax/Meta-Memory — ~16 stars, last commit unknown (10 commits)","storage":"hybrid","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction","unit":"scene-graph","governance":"inspectable","conflict":"append","description":"Formalises spatial localisation question answering (SLQA). High-density memory over 3D scene graphs (semantic entities + metric coordinates); LLM agent performs joint retrieval over both modalities for multi-hop location queries. Introduces SpaceLocQA benchmark.","cited_cells":29,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/meta-memory--arxiv-2509-20754"},{"id":"microsoft-agent-governance-toolkit--gh-microsoft-agent-governance-toolkit","name":"Microsoft Agent Governance Toolkit","url":"https://github.com/microsoft/agent-governance-toolkit","section":"Memory governance, privacy & safety","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Cross-Model Verification Kernel + Agent OS","license":"MIT","latest_release":"v4.1.0 (2026-06-11)","deployment":"library/SDK integration; multi-language (Python TypeScript Rust Go .NET); LangChain/CrewAI/Google ADK integrations","pricing":"free / open source (MIT)","mcp_support":"Yes — MCP Security Scanner (tool poisoning, typosquatting, hidden instructions); A2A/MCP Protocol Bridge in Agent Mesh","a2a_support":"Yes — Agent Mesh provides A2A/MCP/IATP protocol bridges (per agentmesh-platform docs); zero-trust identity + trust scoring","backend_storage":"custom (Microsoft 365)","llm_lock":"","github":"microsoft/agent-governance-toolkit — ~6.1k stars","storage":"append-only","retrieval":"agentic","persistence":"session","update":"read-only","unit":"policy","governance":"auditable","conflict":"llm-arbitrate","description":"Cross-Model Verification Kernel (CMVK) requires majority-voting agreement across multiple model calls before a memory-influenced action. Agent OS package intercepts every action (memory included) at sub-millisecond latency. MIT licensed.","cited_cells":53,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/microsoft-agent-governance-toolkit--gh-microsoft-agent-governance-toolkit"},{"id":"microsoft-copilot-memory--learn-microsoft-com","name":"Microsoft Copilot Memory","url":"https://learn.microsoft.com/en-us/copilot/microsoft-365/copilot-personalization-memory","section":"Platform-provider memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Org / tenant-scoped (Exchange-backed)","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free + paid","mcp_support":"via official adapter — Copilot Studio supports MCP","a2a_support":"supported (Copilot Studio A2A integration)","backend_storage":"custom (Microsoft-managed)","llm_lock":"","github":"","storage":"relational","retrieval":"injection","persistence":"long-term","update":"extraction","unit":"fact","governance":"auditable","conflict":"overwrite","description":"Microsoft 365 Copilot memory (saved memories, chat-history details, custom instructions) stored in a hidden folder of the user's Exchange mailbox under mailbox compliance; on by default, admin-controllable; available to Copilot Chat users with or without an M365 Copilot licence. Docs still label it preview.","cited_cells":68,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/microsoft-copilot-memory--learn-microsoft-com"},{"id":"microsoft-edge-copilot-mode-journeys--blogs-windows-com","name":"Microsoft Edge Copilot Mode + Journeys (now Copilot in Edge)","url":"https://blogs.windows.com/msedgedev/2026/05/13/new-updates-to-edge-across-desktop-and-mobile/","section":"Browser-agent memory","tier":1,"status":"renamed","last_verified_at":"2026-10-29","type":"Tab-context + Journeys session-grouping","license":"","latest_release":"","deployment":"Managed-only (Edge browser with Microsoft account)","pricing":"Free in Edge (Journeys free on desktop in English markets); Browse with Copilot requires Microsoft 365 Premium (US); M365 Personal/Family/Premium get extended usage","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"custom (Microsoft-managed)","llm_lock":"","github":"","storage":"kv|file","retrieval":"injection|similarity","persistence":"cross-session","update":"consolidation|append-only","unit":"episode|summary","governance":"user-controllable","conflict":"none","description":"Copilot Mode retired 2026-05-13; its features now ship as Copilot in Edge. Journeys clusters browsing history into topic cards on the new tab page (desktop, and US mobile); Copilot long-term memory can reference past chats and browsing history with user controls; agentic 'Browse with Copilot' (formerly Copilot Actions) for US Microsoft 365 Premium.","cited_cells":58,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/microsoft-edge-copilot-mode-journeys--blogs-windows-com"},{"id":"microsoft-foundry-agent-service-memory--devblogs-microsoft-com","name":"Microsoft Foundry Agent Service Memory","url":"https://devblogs.microsoft.com/foundry/introducing-memory-in-foundry-agent-service/","section":"Platform-provider memory","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Product","license":"proprietary","latest_release":"searched not found","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"","storage":"proprietary|vector","retrieval":"hybrid|injection","persistence":"long-term","update":"extraction|consolidation","unit":"profile|summary|fact","governance":"n/a","conflict":"llm-arbitrate","description":"Fully managed long-term memory store built into Azure/Microsoft Foundry Agent Service that extracts user-profile facts and chat summaries and retrieves them across sessions — Microsoft's peer to AgentCore Memory and Vertex Memory Bank. Traction: Major-vendor launch (Microsoft Ignite 2025); InfoQ coverage Feb 2026","cited_cells":11,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/microsoft-foundry-agent-service-memory--devblogs-microsoft-com"},{"id":"microsoft-recall--microsoft-com","name":"Microsoft Recall","url":"https://learn.microsoft.com/en-us/windows/apps/develop/windows-integration/recall/","section":"Personal AI / PKM / lifelogging memory","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"OS-level screen-recall timeline","license":"","latest_release":"","deployment":"On-device (snapshots stored locally on Copilot+ PCs)","pricing":"Closed","mcp_support":"searched not found","a2a_support":"not supported","backend_storage":"local SQLite (encrypted)","llm_lock":"","github":"","storage":"file|vector","retrieval":"similarity|extraction-pull","persistence":"long-term","update":"append-only","unit":"episode|document","governance":"user-controllable","conflict":"overwrite","description":"Windows-native screen-recording memory. Continuously indexes activity across apps, documents, browsing; timeline UI for revisiting work. Distinct from Copilot Memory.","cited_cells":41,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/microsoft-recall--microsoft-com"},{"id":"mistral-agents-api--docs-mistral-ai","name":"Mistral Agents API","url":"https://docs.mistral.ai/studio/agents/introduction","section":"Platform-provider memory","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Platform-managed stateful conversation memory","license":"","latest_release":"","deployment":"Managed cloud (la Plateforme) + self-hosted Docker on-prem + Mistral Forge enterprise","pricing":"API pay-per-token (e.g. Mistral Large $0.5 in / $1.5 out per 1M tokens; batch -50%); enterprise custom","mcp_support":"yes — MCP servers registered as managed Connectors usable as tools in agents/conversations","a2a_support":"searched not found","backend_storage":"custom (Mistral-managed)","llm_lock":"","github":"","storage":"kv","retrieval":"injection","persistence":"cross-session","update":"append-only","unit":"episode","governance":"opaque","conflict":"none","description":"Mistral's Agents API provides persistent conversation state across turns and sessions; runtime maintains branching conversation trees. Memory is implicit in the stateful session model rather than a separate store — agents can branch, continue prior discussions, and accumulate tool-call history without manual tracking.","cited_cells":42,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mistral-agents-api--docs-mistral-ai"},{"id":"mistral-le-chat-memories--help-mistral-ai","name":"Mistral Le Chat — Memories (now Vibe)","url":"https://help.mistral.ai/en/articles/396497-how-do-you-handle-my-data-when-using-the-memories-feature","section":"Platform-provider memory","tier":1,"status":"renamed","last_verified_at":"2026-10-29","type":"Graph-contextual + explicit + implicit","license":"","latest_release":"","deployment":"Managed-only (Le Chat) + self-hosted (open weights models) + private cloud via enterprise","pricing":"Free; Pro $14.99/mo; Education $5.99/mo; Team $24.99/user/mo; Enterprise custom","mcp_support":"native (first-party) — Le Chat supports MCP connectors","a2a_support":"not supported","backend_storage":"custom","llm_lock":"","github":"","storage":"graph|kv","retrieval":"graph-traversal|injection","persistence":"long-term","update":"extraction|consolidation","unit":"fact","governance":"user-controllable","conflict":"llm-arbitrate","description":"Le Chat is now Vibe; Mistral's help center describes Memories as a Vibe feature (article dated 2026-08-12). Memories in Mistral's consumer assistant: saves details on user request or automatically; user can view/edit/delete or clear all; beta, retained for subscription duration, may be used for training unless opted out.","cited_cells":64,"third_party_cells":11,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mistral-le-chat-memories--help-mistral-ai"},{"id":"mistral-vibe-remote-agents-mistral-medium-3-5--mistral-ai","name":"Mistral Vibe (Remote Agents) + Mistral Medium 3.5","url":"https://mistral.ai/news/vibe-remote-agents-mistral-medium-3-5","section":"Coding-agent memory","tier":1,"status":"active","last_verified_at":"2026-10-23","type":"Cloud-hosted async coding agents with teleportable session state","license":"Apache-2.0 (Vibe CLI); Modified MIT (Medium 3.5 weights)","latest_release":"mistral-vibe v3.0.1 (2026-10-20)","deployment":"Managed cloud (la Plateforme + Le Chat) + NVIDIA NIM containers","pricing":"API: $1.5/M input tokens; $7.5/M output tokens (Medium 3.5); Vibe via Pro/Team/Enterprise plans","mcp_support":"native (first-party — Mistral Agents support MCP)","a2a_support":"searched not found","backend_storage":"custom (Mistral-managed sandboxes)","llm_lock":"","github":"mistralai/mistral-vibe — ~4.9k stars, last commit not confirmed","storage":"kv","retrieval":"injection","persistence":"cross-session","update":"append-only","unit":"episode","governance":"opaque","conflict":"none","description":"Vibe is Mistral's cloud-hosted coding-agent platform: each session runs in an isolated sandbox with task state, history, and approvals. Spawnable from CLI or Le Chat, with parallel execution. Local CLI sessions can be \"teleported\" to the cloud and resumed. Powered by Mistral Medium 3.5 (77.6% SWE-Bench Verified). Plugs into GitHub, Linear/Jira, Sentry, Slack/Teams.","cited_cells":73,"third_party_cells":11,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mistral-vibe-remote-agents-mistral-medium-3-5--mistral-ai"},{"id":"multi-head-rag-mhrag--arxiv-2406-05085","name":"Multi-Head RAG (MHRAG)","url":"https://arxiv.org/abs/2406.05085","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"Attention-head multi-aspect embeddings","license":"","latest_release":"no-release no-release","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"242★ Python","storage":"vector","retrieval":"similarity|attention","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Uses transformer multi-head attention activations (not decoder-layer embeddings) as retrieval keys, yielding multi-aspect embeddings that independently capture different semantic facets for heterogeneous queries.","cited_cells":24,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/multi-head-rag-mhrag--arxiv-2406-05085"},{"id":"mymind--access-mymind-com","name":"Mymind","url":"https://access.mymind.com/pricing","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Tag-free auto-clustering visual mind","license":"","latest_release":"","deployment":"Managed-only","pricing":"Bookmarker $4.99/mo; Student of Life $7.99/mo or $79/yr; Mastermind $12.99/mo or $129/yr; Newton $299/yr (not yet available)","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"extraction","unit":"document","governance":"user-controllable","conflict":"none","description":"No tags, no folders, no manual organisation. Automatic image tagging, OCR search inside screenshots/memes, smart auto-grouping (Smart Spaces), semantic search. Privacy-first.","cited_cells":33,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mymind--access-mymind-com"},{"id":"n8n-ai-agent-memory--docs-n8n-io","name":"n8n AI Agent Memory","url":"https://docs.n8n.io/advanced-ai/examples/understand-memory/","section":"Framework-embedded memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Pluggable buffer + Postgres/Redis + vector","license":"Sustainable Use License + n8n Enterprise License (fair-code, source-available; not OSI)","latest_release":"n8n@2.44.4 (2026-10-19)","deployment":"Self-hosted (Docker / npm); cloud via n8n.cloud SaaS","pricing":"Free self-hosted Community Edition; Cloud: Starter €20/mo; Pro €50/mo; Business $800/mo (40k executions)","mcp_support":"via official adapter — n8n MCP node","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"any (BYO via LLM nodes)","github":"n8n-io/n8n — ~208k stars, last release 2026-10","storage":"vector|relational","retrieval":"similarity|exact-match","persistence":"cross-session","update":"append-only","unit":"episode|document","governance":"inspectable","conflict":"append-only","description":"Low-code workflow platform. AI Agent node ships with pluggable memory sub-graph: Simple Memory for development; Postgres/Redis for production; vector store nodes (Qdrant, Pinecone, MongoDB Atlas) for semantic recall.","cited_cells":52,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/n8n-ai-agent-memory--docs-n8n-io"},{"id":"nabla-navina--nabla-com","name":"Nabla + Navina","url":"https://www.nabla.com/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Ambient + historical real-time reconciliation","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free trial; pricing via sales (none published)","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|relational","retrieval":"similarity|injection","persistence":"long-term","update":"extraction|consolidation","unit":"episode|document","governance":"auditable","conflict":"editor-in-the-loop","description":"Ambient scribe + longitudinal clinical intelligence (strategic partnership Jul 2025). Live encounter transcript reconciled against historical labs/imaging/notes; condition-suspecting + care-gap engine runs against the reconciled record.","cited_cells":38,"third_party_cells":8,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/nabla-navina--nabla-com"},{"id":"naver-hyperclova-x-think--navercorp-com","name":"NAVER HyperCLOVA X THINK","url":"https://navercorp.com/en/tech/hyperclovax","section":"Dedicated memory layers","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Korean sovereign 128K reasoning LLM","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"not documented publicly","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"","storage":"parametric|kv-cache","retrieval":"attention|parametric-recall","persistence":"session","update":"read-only","unit":"kv-token","governance":"opaque","conflict":"n/a","description":"Reasoning-focused successor to HyperCLOVA X. Pre-trained on ~6T high-quality Korean + English tokens. 128K context via rotary position embeddings + grouped-query attention. Sovereign reasoning LLM for Korean enterprises and government.","cited_cells":33,"third_party_cells":14,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/naver-hyperclova-x-think--navercorp-com"},{"id":"neo4j-agent-memory-neo4j-labs--gh-neo4j-labs-agent-memory","name":"Neo4j Agent Memory (Neo4j Labs)","url":"https://github.com/neo4j-labs/agent-memory","section":"Dedicated memory layers","tier":3,"status":"active","last_verified_at":"2026-10-29","type":"Framework","license":"Apache-2.0","latest_release":"last commit 2026-10-22","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/neo4j-labs/agent-memory (592★)","storage":"graph|vector","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction|consolidation","unit":"entity|episode|trajectory","governance":"inspectable|auditable","conflict":"none","description":"Neo4j Labs' graph-native agent memory library (Python and TypeScript) that stores conversations, builds knowledge/context graphs and records agent reasoning traces to learn from. Also available as NAMS, a hosted memory service (memory.neo4jlabs.com) with server-side extraction and no database to run, alongside the self-hosted Neo4j/Aura backend. Traction: Official vendor-lab project; packages published on PyPI and npm","cited_cells":16,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/neo4j-agent-memory-neo4j-labs--gh-neo4j-labs-agent-memory"},{"id":"neocognition--prnewswire-com","name":"NeoCognition","url":"https://www.prnewswire.com/news-releases/neocognition-emerges-from-stealth-with-40-million-seed-round-to-advance-specialized-intelligence-and-expert-agents-302749108.html","section":"Dedicated memory layers","tier":1,"status":"active","last_verified_at":"2026-10-28","type":"World-model / experiential self-learning memory","license":"","latest_release":"","deployment":"searched not found","pricing":"searched not found","mcp_support":"not documented publicly","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"","storage":"hybrid","retrieval":"agentic","persistence":"long-term","update":"extraction","unit":"episode|fact","governance":"opaque","conflict":"none","description":"Builds agents that learn from experience by constructing a \"world model\" of the micro-environment they operate in. Specialise over time through Agentic RL loop rather than relying on static prompt context. Targets the ~50% task-completion failure rate in current agents.","cited_cells":29,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/neocognition--prnewswire-com"},{"id":"nomi-ai--nomi-ai","name":"Nomi AI","url":"https://nomi.ai","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Mind Map 2.0 — navigable memory graph","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free; Pro ~$15.99/mo, $99.99/yr or $299.99 lifetime","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph","retrieval":"graph-traversal","persistence":"lifelong","update":"extraction","unit":"fact|episode","governance":"inspectable","conflict":"append","description":"AI companion built around long-term memory as primary differentiator. Mind Map 2.0 gives users a navigable graph of what their Nomi knows about them and how memories connect — the most transparent memory model in the companion segment. Users can edit/delete nodes.","cited_cells":34,"third_party_cells":12,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/nomi-ai--nomi-ai"},{"id":"notion-ai--notion-com","name":"Notion AI","url":"https://www.notion.com/product/ai","section":"Personal AI / PKM / lifelogging memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Workspace-scoped context + custom instructions","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free + paid","mcp_support":"yes — official hosted Remote Notion MCP (OAuth); local makenotion/notion-mcp-server no longer maintained","a2a_support":"searched not found","backend_storage":"custom (Notion-managed)","llm_lock":"","github":"","storage":"vector|relational","retrieval":"similarity|injection","persistence":"long-term","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"n/a","description":"Notion's first-party AI. Reads up to 50 pages per query (as of Jan 2026); Custom Instructions persist brand voice / team structure / preferences. Context follows you across model swaps. Notably not persistent cross-session learning — it organises and summarises what's in Notion rather than learning from interactions.","cited_cells":45,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/notion-ai--notion-com"},{"id":"nvidia-inference-context-memory-storage-platform--nvidianews-nvidia-com","name":"NVIDIA Inference Context Memory Storage Platform (now NVIDIA CMX Context Memory Storage Platform)","url":"https://www.nvidia.com/en-us/data-center/ai-storage/cmx/","section":"Dedicated memory layers","tier":2,"status":"renamed","last_verified_at":"2026-10-23","type":"Hardware-accelerated cluster KV-cache storage tier","license":"proprietary (hardware/DOCA) + Dynamo Apache-2.0","latest_release":"","deployment":"On-prem rack-scale (NVIDIA Rubin cluster + partner storage); managed via DOCA","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"custom (BlueField-4 + partner SSD/NVMe arrays)","llm_lock":"","github":"","storage":"kv","retrieval":"similarity","persistence":"session","update":"replacement","unit":"token","governance":"opaque","conflict":"n/a","description":"Renamed NVIDIA CMX Context Memory Storage Platform (2026). A BlueField-4 (STX) powered \"AI-native context tier\" that extends GPU memory with a shared pod-level KV-cache tier for long-context, multi-turn agentic inference, with Dynamo routing and DOCA Memos SDK managing cache across compute and storage nodes. Announced CES Jan 2026; offered via storage partners (Dell, HPE, IBM, NetApp, VAST, WEKA, etc.).","cited_cells":42,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/nvidia-inference-context-memory-storage-platform--nvidianews-nvidia-com"},{"id":"nvidia-isaac-gr00t-n1-7--arxiv-2503-14734","name":"NVIDIA Isaac GR00T N1.7","url":"https://arxiv.org/abs/2503.14734","section":"Vertical / domain-specific AI memory","tier":3,"status":"active","last_verified_at":"2026-10-29","type":"Slow VLM planner + fast diffusion action model","license":"Apache-2.0 (code); model weights under NVIDIA Open Model License","latest_release":"n1.7-release 2026-04-18","deployment":"Open-source (Apache-2.0); self-hosted","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"NVIDIA/Isaac-GR00T — ~8.2k stars, last release 2026-04","storage":"kv-cache","retrieval":"attention","persistence":"session","update":"read-only","unit":"trajectory","governance":"inspectable","conflict":"n/a","description":"Dual-system (VLM \"System 2\" + Diffusion Transformer \"System 1\") jointly trained end-to-end; planner outputs persist as conditioning context across action generation. N1.7 adopts relative end-effector action space — incremental spatial memory via positional history. N1.7 is now a General Availability release with full support and stability guarantees.","cited_cells":33,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/nvidia-isaac-gr00t-n1-7--arxiv-2503-14734"},{"id":"nvidia-remembr--arxiv-2409-13682","name":"NVIDIA ReMEmbR","url":"https://arxiv.org/abs/2409.13682","section":"Vertical / domain-specific AI memory","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Spatio-temporal episodic store (vector DB + coordinates)","license":"Custom","latest_release":"no releases","deployment":"Open-source (Custom NVIDIA license)","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"NVIDIA-AI-IOT/remembr — ~359 stars, 13 commits","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"append-only","unit":"episode","governance":"inspectable","conflict":"append","description":"Builds long-horizon memory by captioning video segments with VILA, storing captions with timestamps + 3D position coordinates in MilvusDB. At query time, LLM iterates retrieval across text, time, and position modalities. Deployed on Nova Carter robot (Jetson Orin).","cited_cells":33,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/nvidia-remembr--arxiv-2409-13682"},{"id":"nyne--nyne-ai","name":"Nyne","url":"https://nyne.ai","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-29","type":"Human-profile / people-graph memory","license":"","latest_release":"","deployment":"Cloud API (agents deployed across web to build people graph; no self-hosted option found)","pricing":"Not public — API access by request","mcp_support":"not documented publicly","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph","retrieval":"graph-traversal|injection","persistence":"long-term","update":"extraction","unit":"profile|fact","governance":"opaque","conflict":"none","description":"Now positioned as 'The data factory for AI': connects AI labs with licensed data and an expert network, built on an identity graph of ~2.2B people and 380M businesses; people-data APIs (identity resolution, enrichment, search) for applications and agents remain available by request.","cited_cells":29,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/nyne--nyne-ai"},{"id":"o-mem--gh-oppo-personalai-o-mem","name":"O-Mem","url":"https://github.com/OPPO-PersonalAI/O-Mem","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Omni / self-evolving memory","license":"Apache-2.0","latest_release":"no releases","deployment":"Self-hosted (research code; no cloud service)","pricing":"not applicable — academic research code (no commercial pricing)","mcp_support":"not documented publicly","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"OPPO-PersonalAI/O-Mem — ~79 stars, last commit n/a","storage":"vector|graph","retrieval":"similarity|graph-traversal","persistence":"lifelong","update":"extraction|consolidation","unit":"fact|episode","governance":"opaque","conflict":"llm-arbitrate","description":"Omni memory system for personalised, long-horizon, self-evolving agents. From OPPO PersonalAI.","cited_cells":36,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/o-mem--gh-oppo-personalai-o-mem"},{"id":"obsidian-smart-connections--gh-brianpetro-obsidian-smart-connections","name":"Obsidian Smart Connections","url":"https://github.com/brianpetro/obsidian-smart-connections","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Local-first semantic search over notes","license":"proprietary (Smart Plugins License, source-available)","latest_release":"v4.7.2 (2026-08-06)","deployment":"Both (local-first plugin; optional cloud LLM APIs)","pricing":"Free core plugin; Pro (all-access) subscription — 14-day free trial then paid","mcp_support":"","a2a_support":"","backend_storage":"file (Obsidian vault)","llm_lock":"","github":"brianpetro/obsidian-smart-connections — ~5.5k stars, last commit 2026-08","storage":"vector|file","retrieval":"similarity","persistence":"long-term","update":"overwrite","unit":"document|chunk","governance":"user-controllable","conflict":"overwrite","description":"Obsidian community plugin. Generates embeddings locally with a bundled model (TaylorAI/bge-micro-v2, 384-dim); chat with notes + see semantically related links. No API key required for local; supports 100+ cloud models optionally. Privacy-first.","cited_cells":40,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/obsidian-smart-connections--gh-brianpetro-obsidian-smart-connections"},{"id":"official-mcp-memory-server--gh-modelcontextprotocol-servers","name":"Official MCP Memory server","url":"https://github.com/modelcontextprotocol/servers/tree/main/src/memory","section":"Claude Code memory mechanisms","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"KG-based persistent memory MCP","license":"Apache-2.0 AND MIT (new contributions Apache-2.0; legacy code MIT; docs CC-BY-4.0)","latest_release":"2026.8.31 (2026-08-31) — @modelcontextprotocol/server-memory@2026.8.31","deployment":"local (Node.js MCP server; local JSON file)","pricing":"free / open source (Apache-2.0 / MIT)","mcp_support":"native (first-party) — reference implementation","a2a_support":"not supported","backend_storage":"file (JSON)","llm_lock":"","github":"modelcontextprotocol/servers — ~90.0k stars, last release 2026-08","storage":"graph","retrieval":"graph-traversal","persistence":"cross-session","update":"overwrite|append-only","unit":"fact","governance":"inspectable","conflict":"llm-arbitrate","description":"Knowledge-graph-based persistent memory MCP server — the canonical reference implementation. npx -y @modelcontextprotocol/server-memory ; configurable via MEMORY_FILE_PATH .","cited_cells":68,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/official-mcp-memory-server--gh-modelcontextprotocol-servers"},{"id":"omega--omegamax-co","name":"OMEGA","url":"https://omegamax.co/","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Memory engine + 50ms retrieval","license":"Apache-2.0","latest_release":"v1.5.20 (2026-10-08)","deployment":"Local-only (local ONNX embeddings + SQLite; macOS + Linux; Windows via WSL)","pricing":"Core free (Apache-2.0); Pro $19/mo (multi-agent coordination, LLM routing, knowledge base, encrypted sync)","mcp_support":"yes — runs as MCP server subprocess; works with Claude Code, Cursor, Windsurf, Cline etc.","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"omega-memory/omega-memory — ~224 stars","storage":"hybrid","retrieval":"similarity","persistence":"long-term","update":"extraction","unit":"fact","governance":"user-controllable","conflict":"none","description":"Local context/memory engine for coding agents: stores decisions, docs and agent history on-machine (SQLite + local ONNX embeddings) and serves a ranked slice per task over MCP. Open-core Apache-2.0 core with paid Pro tier.","cited_cells":32,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/omega--omegamax-co"},{"id":"omi--omi-me","name":"Omi","url":"https://www.omi.me","section":"Voice-first / wearable AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Open-source wearable orb + memory graph","license":"MIT","latest_release":"v0.12.522+12522-macos (2026-10-22)","deployment":"Both","pricing":"Hardware-bundled","mcp_support":"yes — repo ships an MCP server (mcp/ folder)","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"BasedHardware/omi — ~13.2k stars, last commit 2026-10","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"lifelong","update":"append-only|consolidation","unit":"episode|fact","governance":"user-controllable","conflict":"append","description":"Wearable orb (necklace or clip) with open-source hardware + firmware. Continuous audio capture; real-time GPT-4o transcription; searchable memory graph. Supports app screen capture too. Hybrid privacy: cloud by default (SOC 2/HIPAA/AES-256), full local/self-hosted available under MIT.","cited_cells":56,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/omi--omi-me"},{"id":"openai-agents-sdk-memory--openai-github-io","name":"OpenAI Agents SDK Memory","url":"https://openai.github.io/openai-agents-python/ref/memory/","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Sessions + RunContextWrapper + categorised long-term","license":"MIT","latest_release":"v0.24.0 (2026-10-15)","deployment":"Both (OSS library self-hosted; relies on OpenAI or Azure OpenAI API endpoints; no dedicated SDK hosting)","pricing":"SDK OSS free; pay for OpenAI API model usage at standard token pricing (e.g. GPT-5.4: $2.50/$15 per M tokens)","mcp_support":"native (first-party) — Agents SDK MCP support","a2a_support":"searched not found","backend_storage":"custom (OpenAI-managed)","llm_lock":"single-vendor (OpenAI; OpenAI-Compatible adapters possible)","github":"openai/openai-agents-python — ~28.6k stars","storage":"kv|vector","retrieval":"injection|similarity","persistence":"cross-session","update":"overwrite|extraction","unit":"episode|fact","governance":"inspectable","conflict":"overwrite","description":"OpenAI's agent SDK. Sessions for working context within an agent loop; RunContextWrapper for structured persistent state. April 2026 update added categorised long-term memory: project / user / policy. Long-term tier typically backed by external vector DBs (Pinecone, etc.).","cited_cells":51,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/openai-agents-sdk-memory--openai-github-io"},{"id":"openai-chatgpt-memory--openai-com","name":"OpenAI ChatGPT Memory","url":"https://openai.com/index/memory-and-new-controls-for-chatgpt/","section":"Platform-provider memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"User-scoped persistent facts","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free + paid","mcp_support":"native (first-party) — ChatGPT apps support MCP connectors","a2a_support":"not supported","backend_storage":"custom (OpenAI-managed)","llm_lock":"","github":"","storage":"kv","retrieval":"injection","persistence":"long-term","update":"extraction","unit":"fact","governance":"user-controllable","conflict":"llm-arbitrate","description":"First-party ChatGPT memory combining saved memories, reference to past chats, custom instructions and connected apps; 'improved memory' keeps an evolving summary (legacy list-style saved memories still selectable), with review/correct/delete controls and temporary chats.","cited_cells":70,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/openai-chatgpt-memory--openai-com"},{"id":"openai-codex-cloud-agent--developers-openai-com","name":"OpenAI Codex (cloud agent)","url":"https://developers.openai.com/codex/memories","section":"Coding-agent memory","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"File-backed ~/.codex/memories + Chronicle","license":"Apache-2.0 (Codex CLI; cloud agent proprietary)","latest_release":"0.165.0 (2026-10-27) (Codex CLI)","deployment":"Both (cloud agent via ChatGPT platform + CLI tool local)","pricing":"Included in ChatGPT Plus/Pro/Business/Edu/Enterprise plans (memory feature)","mcp_support":"native (first-party)","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"openai/codex — ~128k stars, last commit 2026-10","storage":"file","retrieval":"injection|extraction-pull","persistence":"long-term","update":"consolidation|append-only","unit":"summary|episode|fact","governance":"inspectable","conflict":"none","description":"Memory under ~/.codex/memories/ : summaries, durable entries, recent-input logs, supporting evidence — generated in background after sessions go idle. Chronicle (opt-in macOS preview) uses screen-capture to seed memory. Off by default; opt-in via config.toml .","cited_cells":48,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/openai-codex-cloud-agent--developers-openai-com"},{"id":"openai-codex-memories-chronicle--developers-openai-com","name":"OpenAI Codex Memories + Chronicle","url":"https://learn.chatgpt.com/docs/customization/memories","section":"Coding-agent memory","tier":3,"status":"active","last_verified_at":"2026-10-29","type":"Product","license":"proprietary","latest_release":"searched not found","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"","storage":"file","retrieval":"injection","persistence":"long-term","update":"extraction|consolidation","unit":"fact|summary","governance":"user-controllable|plaintext","conflict":"none","description":"Codex Memories carries preferences, conventions and pitfalls across Codex threads as local memory files. The Chronicle research preview has been replaced by Computer History, a rebuilt macOS desktop feature (ChatGPT Pro, Business, Enterprise; off by default) that turns interaction events and accessibility text from allowed apps and websites, without screenshots or audio, into summaries, local memory files and a timeline that ChatGPT and Codex can reference. Traction: OpenAI first-party feature for the Codex app; wide press coverage","cited_cells":11,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/openai-codex-memories-chronicle--developers-openai-com"},{"id":"openclaw-memory--docs-openclaw-ai","name":"OpenClaw memory","url":"https://docs.openclaw.ai/concepts/memory","section":"File-backed / editor paradigms","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"MIT","latest_release":"last commit 2026-10-22","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/openclaw/openclaw (392751★)","storage":"file|vector","retrieval":"injection|agentic","persistence":"long-term","update":"agent-controlled|consolidation","unit":"fact|profile|episode","governance":"editable","conflict":"overwrite","description":"Memory system of the viral OpenClaw personal agent: plain Markdown files (USER.md, MEMORY.md, daily memory/YYYY-MM-DD.md notes, DREAMS.md) with memory_search/memory_get tools, automatic pre-compaction flush and a 'dreaming' sweep that distils daily notes into long-term memory; pluggable engines (LanceDB, Honcho, wiki). spawned an ecosystem of memory plugins (memory-lancedb-pro, gbrain, MemClaw)","cited_cells":19,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/openclaw-memory--docs-openclaw-ai"},{"id":"openevidence-deepconsult-visits--openevidence-com","name":"OpenEvidence (DeepConsult + Visits)","url":"https://www.openevidence.com/","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Encounter-grounded evidence synthesis","license":"","latest_release":"","deployment":"Managed-only (web + mobile apps)","pricing":"Free for verified physicians/HCPs; ad-supported","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"session","update":"read-only","unit":"document","governance":"inspectable","conflict":"none","description":"Clinical decision support. Visits (Aug 2025) connects evidence layer to the live encounter — patient-specific context binding distinct from pure literature RAG. DeepConsult agent autonomously cross-references hundreds of peer-reviewed studies.","cited_cells":27,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/openevidence-deepconsult-visits--openevidence-com"},{"id":"openmemory--gh-caviraoss-openmemory","name":"OpenMemory (now LongMemory)","url":"https://github.com/CaviraOSS/LongMemory","section":"Dedicated memory layers","tier":2,"status":"renamed","last_verified_at":"2026-10-29","type":"Cognitive memory engine","license":"Apache-2.0","latest_release":"v1.2.3 (2025-12-12); v1.3.0 beta pre-release (2025-12-20)","deployment":"Self-hosted (local + Docker; no cloud service found)","pricing":"OSS (Apache-2.0 self-hosted); self-hosted cost ~$5-8/mo for 100k memories vs $60-120 for SaaS","mcp_support":"yes — built-in stdio MCP (longmemory mcp) and authenticated Streamable HTTP MCP","a2a_support":"not documented publicly","backend_storage":"hybrid (vector + graph)","llm_lock":"","github":"CaviraOSS/LongMemory (formerly OpenMemory) — ~4.5k stars, last commit 2026-10","storage":"vector|graph","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction","unit":"fact","governance":"user-controllable","conflict":"none","description":"Renamed LongMemory (repo CaviraOSS/LongMemory; packages longmemory / longmemory-sdk, with openmemory-js@2 / openmemory-py@2 as forwarding bridges). Open-source cognitive memory engine for LLMs and agents.","cited_cells":47,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/openmemory--gh-caviraoss-openmemory"},{"id":"openmemory-mcp--mem0-ai","name":"OpenMemory MCP","url":"https://mem0.ai/openmemory","section":"Claude Code memory mechanisms","tier":2,"status":"active","last_verified_at":"2026-10-29","type":"Mem0-hosted MCP memory server (cross-tool memory)","license":"Apache-2.0 (mem0 monorepo)","latest_release":"","deployment":"","pricing":"","mcp_support":"native (first-party)","a2a_support":"not supported","backend_storage":"hybrid","llm_lock":"","github":"mem0ai/mem0 — ~67.3k stars (openmemory/ subdirectory removed from main by Oct 2026; MCP server is hosted)","storage":"vector|relational","retrieval":"similarity","persistence":"long-term","update":"extraction","unit":"fact","governance":"user-controllable","conflict":"llm-arbitrate","description":"Mem0's MCP memory layer for coding agents (Claude, Cursor, VS Code, Windsurf, Codex etc.). As of Oct 2026 it is a hosted service: the MCP server runs at mcp.mem0.ai, needs a Mem0 Platform account (browser sign-in or API key, npx mcp-add installer), and memories live in the Mem0 account, not on the user's machine. The earlier local Docker OpenMemory server (openmemory/ folder) is no longer in the mem0ai/mem0 repo; self-hosting now means the general Mem0 open-source server.","cited_cells":37,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/openmemory-mcp--mem0-ai"},{"id":"openviking-bytedance-volcengine--gh-volcengine-openviking","name":"OpenViking (ByteDance Volcengine)","url":"https://github.com/volcengine/OpenViking","section":"Dedicated memory layers","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"AGPL-3.0 (CLI/examples Apache-2.0)","latest_release":"last commit 2026-10-22","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/volcengine/OpenViking (40582★)","storage":"file|vector","retrieval":"similarity|agentic|exact-match","persistence":"long-term","update":"extraction|consolidation","unit":"file|summary|skill","governance":"editable|inspectable","conflict":"n/a","description":"Open-source 'context database' from ByteDance's Volcengine that unifies agent memory, RAG knowledge and skills behind a virtual filesystem (viking:// URIs) with tiered loading and self-evolving context. #6 on GitHub agent-memory topic; listed #3 in TeleAI Awesome-Agent-Memory products","cited_cells":17,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/openviking-bytedance-volcengine--gh-volcengine-openviking"},{"id":"opera-ai-formerly-aria--blogs-opera-com","name":"Opera AI (formerly Aria)","url":"https://blogs.opera.com/news/2025/02/aria-gets-memory-feature-new-ai-feature-drop/","section":"Browser-agent memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"User-declared preferences + tab management","license":"","latest_release":"","deployment":"Managed-only (Opera browser)","pricing":"Free (included with Opera browser); Opera One Pro has additional AI credits","mcp_support":"yes (Opera Neon only) — MCP Connector lets external AI clients drive the browser (Mar 2026); simplified connector promised for Opera One/GX","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"kv","retrieval":"injection","persistence":"long-term","update":"overwrite","unit":"fact","governance":"user-controllable","conflict":"none","description":"Memory feature stores user-declared preferences and facts; injected into future prompts. Understands current tab content; manages open tabs contextually. Memory off by default, end-user controlled, stored alongside the prompt rather than the LLM.","cited_cells":38,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/opera-ai-formerly-aria--blogs-opera-com"},{"id":"otter-ai--otter-ai","name":"Otter.ai","url":"https://otter.ai","section":"Voice-first / wearable AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Meeting bot + browser extension + app","license":"","latest_release":"","deployment":"Managed-only","pricing":"Basic free; Pro $16.99/user/mo monthly or $8.33 annual; Business $30 list ($19.99 annual); Enterprise custom","mcp_support":"yes — first-party Otter MCP server (listed in Basic/Pro features); API & webhooks on Enterprise","a2a_support":"not supported","backend_storage":"custom","llm_lock":"","github":"","storage":"file|vector","retrieval":"similarity|extraction-pull","persistence":"long-term","update":"append-only","unit":"episode|document","governance":"inspectable","conflict":"append","description":"Joins Zoom / Meet / Teams as a meeting bot. Real-time transcription, speaker identification, searchable archive across past meetings. SSO + admin controls at Business tier.","cited_cells":49,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/otter-ai--otter-ai"},{"id":"owasp-agent-memory-guard--owasp-org","name":"OWASP Agent Memory Guard","url":"https://owasp.org/www-project-agent-memory-guard/","section":"Memory governance, privacy & safety","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Declarative YAML policies + SHA-256 baselines","license":"Apache-2.0","latest_release":"v0.3.1 (2026-08-25)","deployment":"local deployment; Python; run against local memory stores","pricing":"free / open source","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"OWASP/www-project-agent-memory-guard — ~185 stars, last commit 2026-08","storage":"file","retrieval":"exact-match","persistence":"long-term","update":"read-only","unit":"file","governance":"auditable","conflict":"n/a","description":"Open-source runtime defense. Enforces YAML policies on every memory read/write. SHA-256 baselines detect tampering, injection, sensitive-data leakage, protected-key modification, rapid-change anomalies. Forensic snapshots + rollback. Reference impl for OWASP ASI06 (Memory Poisoning).","cited_cells":40,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/owasp-agent-memory-guard--owasp-org"},{"id":"pathrag--arxiv-2502-14902","name":"PathRAG","url":"https://arxiv.org/abs/2502.14902","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"Path-pruned graph RAG","license":"MIT","latest_release":"n/a","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"BUPT-GAMMA/PathRAG — ~376 stars","storage":"graph","retrieval":"graph-traversal","persistence":"long-term","update":"accretion","unit":"fact","governance":"plaintext","conflict":"n/a","description":"Retrieves over an indexed graph by pruning candidate relational paths — reduces noise in graph-RAG vs flat path expansion.","cited_cells":10,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/pathrag--arxiv-2502-14902"},{"id":"pathrag--gh-bupt-gamma-pathrag","name":"PathRAG","url":"https://github.com/BUPT-GAMMA/PathRAG","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"Path-pruned graph retrieval","license":"MIT","latest_release":"no releases","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"BUPT-GAMMA/PathRAG — ~376 stars, last commit not confirmed","storage":"graph","retrieval":"graph-traversal|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Retrieves key relational paths from knowledge graph; flow-based pruning + reliability scoring; path-based prompting. Outperforms baselines on six datasets. arxiv Feb 2025.","cited_cells":27,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/pathrag--gh-bupt-gamma-pathrag"},{"id":"perplexity-brain--perplexity-ai","name":"Perplexity Brain","url":"https://www.perplexity.ai/","section":"Platform-provider memory","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Product","license":"proprietary","latest_release":"searched not found","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"","storage":"proprietary","retrieval":"agentic","persistence":"long-term","update":"extraction","unit":"fact","governance":"inspectable","conflict":"n/a","description":"Topic-organised shared memory layer under Perplexity Search, the Computer agent and possibly Comet; stores projects, decisions and sources, retrieves selectively per task, and exposes topics plus a 3D connection map to the user. Traction: Major-vendor announcement; Perplexity claims +25% correctness, +16% recall, -13% cost on context-dependent tasks","cited_cells":10,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/perplexity-brain--perplexity-ai"},{"id":"perplexity-comet--perplexity-ai","name":"Perplexity Comet","url":"https://www.perplexity.ai/comet","section":"Browser-agent memory","tier":1,"status":"active","last_verified_at":"2026-10-23","type":"Conversation across pages + browser-history memory","license":"","latest_release":"","deployment":"Managed (desktop Mac/Windows + iOS/Android apps)","pricing":"Free (since Oct 2025); optional Comet Plus $5/mo; Perplexity Pro $20/mo / Max $200/mo","mcp_support":"Comet Assistant supports MCP servers via comet://settings/connectors","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"file|vector","retrieval":"similarity|extraction-pull","persistence":"long-term","update":"append-only","unit":"episode|document","governance":"opaque","conflict":"none","description":"AI-native Chromium-based browser. Assistant operates across tabs and sessions; persistent memory of preferences; multi-step agentic tasks without leaving the browser. Memory is opt-in and server-side; enterprise memory tier separately available.","cited_cells":66,"third_party_cells":8,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/perplexity-comet--perplexity-ai"},{"id":"perplexity-memory--perplexity-ai","name":"Perplexity Memory","url":"https://www.perplexity.ai/help-center/en/articles/10968016-memory","section":"Platform-provider memory","tier":1,"status":"active","last_verified_at":"2026-10-23","type":"Auto-distillation across model swaps","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free + paid","mcp_support":"yes — first-party Perplexity MCP server (remote at api.perplexity.ai/mcp; local perplexityai/modelcontextprotocol, npm @perplexity-ai/mcp-server) exposing search/ask/research/reason","a2a_support":"not supported","backend_storage":"custom","llm_lock":"","github":"","storage":"kv","retrieval":"injection","persistence":"long-term","update":"extraction","unit":"fact","governance":"user-controllable","conflict":"llm-arbitrate","description":"Stores preferences and details (and optionally search history) to personalize answers, citing which memory was used; user controls in Settings → Memory. 'Brain' adds structured self-improving memory for Max subscribers and shared project Brains in Computer; Enterprise memories are org-owned and admin-managed.","cited_cells":49,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/perplexity-memory--perplexity-ai"},{"id":"personal-ai--personal-ai","name":"Personal.ai","url":"https://www.personal.ai/","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"active","last_verified_at":"2026-10-29","type":"Personal AI with unique memory","license":"","latest_release":"","deployment":"Deployed on the operator's own network (carrier on-prem)","pricing":"No public pricing — sales-led single agreement per operator (deployment, licensing, support)","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"lifelong","update":"extraction","unit":"fact|episode","governance":"user-controllable","conflict":"none","description":"AI memory platform sold to telecom carriers: memory, identity and on-network inference that turn each subscriber identity on the network into an agent (products: My AI, Persona Studio, Memory Core; MODEL-4).","cited_cells":36,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/personal-ai--personal-ai"},{"id":"pieces-for-developers-ltm--pieces-app","name":"Pieces for Developers (LTM)","url":"https://pieces.app/","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"On-device long-term memory for IDEs","license":"searched not found","latest_release":"searched not found","deployment":"Local + hybrid cloud sync","pricing":"searched not found","mcp_support":"native (PiecesOS MCP server)","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"pieces-app/cli-agent — ~113 stars (CLI only; PiecesOS core not on GitHub), last commit n/a","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"accretion","unit":"episode","governance":"plaintext","conflict":"n/a","description":"Long-term memory engine for developers — captures everything the user does (browser, IDE, terminal) on-device, exposes via MCP for any IDE/agent.","cited_cells":13,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/pieces-for-developers-ltm--pieces-app"},{"id":"plaud-notepin--plaud-ai","name":"Plaud NotePin","url":"https://www.plaud.ai","section":"Voice-first / wearable AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Magnetic clip-on + AI transcription","license":"","latest_release":"","deployment":"Managed-only","pricing":"NotePin $159 / NotePin S $179 / Note Pro $189; Starter plan free (300 min/mo); Pro $17.99/mo or $99.99/yr (1,200 min); Unlimited $29.99/mo or $239.99/yr","mcp_support":"yes — first-party MCP & CLI (docs.plaud.ai) plus device SDK/API","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"file","retrieval":"extraction-pull","persistence":"long-term","update":"append-only","unit":"episode|document","governance":"user-controllable","conflict":"append","description":"Magnetic clip-on wearable. Records in-person conversations and calls; AI summaries, action items, searchable transcripts. AES-256 + TLS 1.2 in transit; doesn't retain audio after transcription. Also sells Note (card-form recorder) and Note Pro.","cited_cells":37,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/plaud-notepin--plaud-ai"},{"id":"polyai--poly-ai","name":"PolyAI","url":"https://poly.ai","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Runtime context-orchestration framework","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"yes — hosted Builder MCP and Data MCP servers; Agent Studio agents can call external MCP servers","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"kv","retrieval":"injection|agentic","persistence":"session","update":"overwrite","unit":"fact","governance":"inspectable","conflict":"slot-filling","description":"Enterprise voice agents. Context heavily runtime-injected from CRM / telephony / API / user history; passes context to humans on escalation. Strong on plumbing, lighter on autonomous memory accumulation.","cited_cells":31,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/polyai--poly-ai"},{"id":"praison-memory--gh-mervinpraison-praisonai","name":"Praison Memory","url":"https://github.com/MervinPraison/PraisonAI","section":"Framework-embedded memory","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Multi-agent framework memory module","license":"MIT","latest_release":"v4.7.16 (2026-10-19)","deployment":"searched not found","pricing":"free / open source (MIT)","mcp_support":"yes — MCP client (stdio/HTTP/WebSocket/SSE) via MCP() tools","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"MervinPraison/PraisonAI — ~9k stars, last commit 2026-10","storage":"vector","retrieval":"similarity","persistence":"long-term|session","update":"accretion","unit":"episode","governance":"plaintext","conflict":"n/a","description":"PraisonAI multi-agent framework memory: zero-dependency default (memory=True), short/long-term and file memory, DB persistence (Postgres/SQLite/Redis/Mongo etc.), graph memory, and a Claude memory-tool integration.","cited_cells":18,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/praison-memory--gh-mervinpraison-praisonai"},{"id":"pydantic-ai-hindsight--hindsight-vectorize-io","name":"Pydantic-AI + Hindsight","url":"https://hindsight.vectorize.io/blog/2026/03/09/pydantic-ai-persistent-memory","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"External memory bolted on","license":"MIT","latest_release":"hindsight-pydantic-ai v0.4.20 (2026-04-15); pydantic-ai v2.59.0 (2026-10-21)","deployment":"Both","pricing":"OSS","mcp_support":"native (first-party) — Pydantic-AI MCP server/client","a2a_support":"supported — Pydantic-AI A2A","backend_storage":"searched not found","llm_lock":"any (BYO via Anthropic/OpenAI/Google/Groq/Mistral providers)","github":"pydantic/pydantic-ai — ~19.6k stars, last commit 2026-10","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"consolidation","unit":"episode","governance":"opaque","conflict":"none","description":"Pydantic-AI ships no native persistent memory; agents start fresh each run. Long-term memory typically added via third-party integration (e.g. Hindsight) using tools + auto-injected memory instructions.","cited_cells":56,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/pydantic-ai-hindsight--hindsight-vectorize-io"},{"id":"pydantic-ai-memorytool--pydantic-dev","name":"Pydantic-AI MemoryTool","url":"https://pydantic.dev/docs/ai/tools-toolsets/native-tools/","section":"Framework-embedded memory","tier":3,"status":"active","last_verified_at":"2026-10-29","type":"Tool-delegated to model provider","license":"MIT","latest_release":"","deployment":"Both (OSS self-hosted; Pydantic Logfire cloud for observability)","pricing":"OSS free (MIT for pydantic-ai); pay for model provider API usage; Pydantic Logfire observability has separate pricing","mcp_support":"native (first-party)","a2a_support":"supported — Pydantic-AI A2A","backend_storage":"","llm_lock":"single-vendor (Anthropic only for memory tool; multi-vendor for chat)","github":"pydantic/pydantic-ai — ~19.6k stars, last commit n/a","storage":"proprietary","retrieval":"agentic","persistence":"cross-session","update":"agent-controlled","unit":"fact","governance":"opaque","conflict":"none","description":"Pydantic AI Harness ships a Memory capability: a MEMORY.md notebook plus extra Markdown notes, exposed via write/read/delete/search memory tools and auto-injected into context; backends InMemory, File, SQLite, Postgres with per-user namespaces. Provider memory tools (e.g. Anthropic) remain available as builtin tools.","cited_cells":52,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/pydantic-ai-memorytool--pydantic-dev"},{"id":"quizlet-ai-study-tools-magic-notes--quizlet-com","name":"Quizlet — AI Study Tools / Magic Notes","url":"https://quizlet.com/features/ai-study-tools","section":"Vertical / domain-specific AI memory","tier":1,"status":"dead","last_verified_at":"2026-10-22","type":"Per-set Memory Score (SRS) + Magic Notes generation","license":"","latest_release":"","deployment":"iOS + Android + Web","pricing":"Free; Quizlet Plus $19.99/yr","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"relational","retrieval":"injection","persistence":"lifelong","update":"overwrite","unit":"profile|fact","governance":"inspectable","conflict":"llm-arbitrate","description":"Tracks per-card answer history within a study set; computes a Memory Score for SRS scheduling. Magic Notes converts uploaded content (notes, PDFs, URLs) into flashcard sets automatically. Q-Chat, the conversational AI layer, was discontinued June 2025. Personalization is scoped to individual study sets, not a unified cross-subject learner model.","cited_cells":28,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/quizlet-ai-study-tools-magic-notes--quizlet-com"},{"id":"raft--arxiv-2403-10131","name":"RAFT","url":"https://arxiv.org/abs/2403.10131","section":"Retrieval-as-memory hybrids","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Retrieval-augmented fine-tuning","license":"Apache-2.0","latest_release":"n/a","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"ShishirPatil/gorilla (raft/ subdir) — ~13.0k stars","storage":"parametric","retrieval":"parametric-recall","persistence":"parametric-permanent","update":"parametric-edit","unit":"document","governance":"inspectable","conflict":"n/a","description":"UC Berkeley. Trains the model to ignore distractor documents and quote the relevant text in chain-of-thought.","cited_cells":27,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/raft--arxiv-2403-10131"},{"id":"raptor--arxiv-2401-18059","name":"RAPTOR","url":"https://arxiv.org/abs/2401.18059","section":"Retrieval-as-memory hybrids","tier":3,"status":"research-code","last_verified_at":"2026-10-22","type":"Recursive abstractive tree RAG","license":"MIT","latest_release":"n/a (no GitHub releases)","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"parthsarthi03/raptor — ~1.8k stars, last commit not confirmed","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"accretion","unit":"chunk","governance":"plaintext","conflict":"n/a","description":"Stanford recursive-summarisation tree over the corpus — clusters and summarises documents into a hierarchical index; retrieves at multiple levels. ICLR 2024.","cited_cells":13,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/raptor--arxiv-2401-18059"},{"id":"raptor--gh-parthsarthi03-raptor","name":"RAPTOR","url":"https://github.com/parthsarthi03/raptor","section":"Retrieval-as-memory hybrids","tier":3,"status":"research-code","last_verified_at":"2026-10-23","type":"Recursive abstractive summary tree","license":"MIT","latest_release":"no releases","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"parthsarthi03/raptor — ~1.8k stars, 20 commits","storage":"vector","retrieval":"similarity|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Recursively embeds, clusters, and summarises chunks into a tree of abstractions. Retrieval traverses the tree at query time. ICLR 2024.","cited_cells":27,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/raptor--gh-parthsarthi03-raptor"},{"id":"ratine--gh-goweft-ratine","name":"Ratine","url":"https://github.com/goweft/ratine","section":"Memory observability & monitoring","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Offline scanner — memory poisoning detector","license":"MIT","latest_release":"v0.1.0 2026-04-02","deployment":"local CLI — Python; pip install; no server required","pricing":"free / open source (MIT); no paid tier","mcp_support":"Yes — built for MCP tool-server ecosystem (sibling tools heddle/loom are MCP runtimes); detects agent memory poisoning across MCP sessions","a2a_support":"no data — searched github.com/goweft/ratine README; single-purpose memory-poisoning detector for MCP tool servers, no A2A protocol","backend_storage":"searched not found","llm_lock":"","github":"1★ Python","storage":"file","retrieval":"extraction-pull","persistence":"cross-session","update":"read-only","unit":"fact","governance":"auditable","conflict":"n/a","description":"Only tool found that scans the persistent memory layer on disk rather than runtime tracing. Detects injected instructions, obfuscated payloads (zero-width Unicode, base64, homoglyphs, hex), C2-pattern URLs, credential leakage. ratine diff compares snapshots for belief drift.","cited_cells":42,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/ratine--gh-goweft-ratine"},{"id":"read-ai--read-ai","name":"Read.ai","url":"https://www.read.ai","section":"Voice-first / wearable AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Meeting bot + analytics + Ada digital twin","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free (5 meetings/mo); Pro $19.75/mo ($15 annual); Enterprise $29.75 ($22.50); Enterprise+ $39.75 ($29.75), per user","mcp_support":"yes — official Read AI MCP server, plus Claude Connector and ChatGPT app","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"file|vector","retrieval":"similarity|extraction-pull","persistence":"long-term","update":"append-only","unit":"episode|document","governance":"inspectable","conflict":"append","description":"Meeting bot + browser extension. Captures audio + video; engagement scoring (talk-time, sentiment, attention); enterprise search across all past meetings. \"Ada\" digital-twin feature builds an AI memory of work patterns.","cited_cells":45,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/read-ai--read-ai"},{"id":"realm--arxiv-2002-08909","name":"REALM","url":"https://arxiv.org/abs/2002.08909","section":"Retrieval-as-memory hybrids","tier":3,"status":"research-code","last_verified_at":"2026-10-23","type":"Retrieval-augmented LM pre-training","license":"Apache-2.0","latest_release":"no releases","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"1.8k★ Python","storage":"vector","retrieval":"similarity","persistence":"parametric-permanent","update":"parametric-edit","unit":"document","governance":"inspectable","conflict":"n/a","description":"Foundational paper. Augments LM pre-training with a latent knowledge retriever — model retrieves and attends over Wikipedia-scale corpus during pre-training, fine-tuning, and inference. Backprop through retrieval over millions of documents. Guu et al., Google, ICML 2020.","cited_cells":25,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/realm--arxiv-2002-08909"},{"id":"redis-agent-memory-server--gh-redis-agent-memory-server","name":"Redis Agent Memory Server","url":"https://github.com/redis/agent-memory-server","section":"Dedicated memory layers","tier":3,"status":"deprecated","last_verified_at":"2026-10-29","type":"Framework","license":"Apache-2.0","latest_release":"last commit 2026-09-15","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/redis/agent-memory-server (323★)","storage":"vector|kv","retrieval":"similarity","persistence":"long-term|session","update":"extraction|consolidation","unit":"fact|episode|summary","governance":"editable","conflict":"llm-arbitrate","description":"Redis's original open-source working + long-term memory server for agents (REST and MCP), with automatic extraction, semantic search and topic/entity tagging; also used to give Snowflake Cortex Agents persistent memory. Superseded by Redis Agent Memory, a managed service on Redis Cloud that is part of Redis Iris; the open-source server now lives in the repo's V0/ folder as a research artifact, not the supported production distribution.","cited_cells":14,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/redis-agent-memory-server--gh-redis-agent-memory-server"},{"id":"reflect--reflect-app","name":"Reflect","url":"https://reflect.app/","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Apple-ecosystem second brain","license":"MIT","latest_release":"","deployment":"Local-first (Markdown files on device; Mac/iPhone/iPad); Classic remains managed","pricing":"Reflect Open: Mac app free; iPhone/iPad paid subscription (App Store); AI via own API key","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"team-reflect/reflect-open — ~1.5k stars, last commit unknown (1,107 commits)","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"overwrite","description":"Daily-notes + backlinks note app now relaunched as Reflect Open: open-source, offline-first, notes stored as local Markdown files synced via iCloud or Git; optional AI with your own API key. Reflect Classic still maintained.","cited_cells":37,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/reflect--reflect-app"},{"id":"relevance-ai-memory--relevanceai-com","name":"Relevance AI Memory","url":"https://relevanceai.com/docs/agent/build-your-agent/memory","section":"Framework-embedded memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Short-term metadata + persistent KV store","license":"","latest_release":"","deployment":"Managed cloud only (SaaS); multi-region on enterprise tier","pricing":"Public pricing page now lists only Enterprise (custom actions/vendor credits; contact sales)","mcp_support":"yes — hosted MCP server at mcp.relevanceai.com exposing agents/tools/knowledge to Claude, ChatGPT, Cursor etc. (OAuth, role-scoped); MCP client for external servers also documented","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"multi-vendor","github":"","storage":"kv","retrieval":"injection|exact-match","persistence":"cross-session","update":"overwrite","unit":"fact","governance":"inspectable","conflict":"overwrite","description":"Two memory scopes: conversation metadata (KV per-conversation) and persistent memory (cross-session). Enabling metadata auto-adds Add/Read conversation metadata tools.","cited_cells":42,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/relevance-ai-memory--relevanceai-com"},{"id":"remnote-ai--remnote-com","name":"RemNote AI","url":"https://www.remnote.com/","section":"Personal AI / PKM / lifelogging memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Outline + bidirectional + spaced repetition","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free; Pro $8/mo ($96/yr); Pro with AI $18/mo ($216/yr); lifetime option","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"graph|relational","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"append-only|overwrite","unit":"fact|episode","governance":"user-controllable","conflict":"overwrite","description":"Hierarchical outline + bidirectional links + spaced-repetition flashcards. AI auto-generates flashcards/quizzes from notes; AI summaries; fill-gap questions.","cited_cells":40,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/remnote-ai--remnote-com"},{"id":"replicant--replicant-com","name":"Replicant","url":"https://www.replicant.com","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Session context + structured handoff","license":"","latest_release":"","deployment":"Managed SaaS; multi-region failover; integrates with CCaaS/CRM/telephony; SOC 2, GDPR, HIPAA, PCI","pricing":"searched not found","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"kv","retrieval":"injection","persistence":"session","update":"overwrite","unit":"fact","governance":"inspectable","conflict":"pii","description":"Enterprise agentic customer-service platform with Conversation Automation (one agent across voice, chat and messaging) and Conversation Intelligence (analysis of human-agent conversations used to replicate workflows). Cross-interaction memory is not documented.","cited_cells":27,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/replicant--replicant-com"},{"id":"replika--replika-ai","name":"Replika","url":"https://replika.ai","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Memory bank + diary entries","license":"","latest_release":"","deployment":"Managed-only","pricing":"Classic: Free/Pro/Ultra/Platinum; Replika 2.0: Plus and Max plans (e.g. ~£59–66/yr in UK, Aug 2026)","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|kv","retrieval":"similarity|injection","persistence":"lifelong","update":"extraction","unit":"episode|fact","governance":"opaque","conflict":"append","description":"Long-running AI companion (Luka, Inc., 2017). Replika 2.0 (2026) is a full rebuild with a new memory architecture, 3D avatars and voice/video calls; users report the new segmented, recency-biased memory losing some prior details. Classic app remains.","cited_cells":34,"third_party_cells":14,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/replika--replika-ai"},{"id":"replit-agent--replit-com","name":"Replit Agent","url":"https://replit.com/products/agent","section":"Coding-agent memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Project-files-as-memory + Temporal orchestration","license":"","latest_release":"","deployment":"Managed-only","pricing":"Starter free; Core $20/mo ($18 annual); Pro $100/mo ($90 annual); Enterprise custom","mcp_support":"yes — Agent connects to listed or custom MCP servers via Integrations (OAuth or custom headers), with one-click install links","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"inspectable","conflict":"none","description":"No documented built-in persistent semantic memory store. Session duration expanded to 200 min (Agent 3); Agent 4 shifts toward persistent runtimes. Temporal used for orchestration durability, not semantic memory. Cross-session memory relies on filesystem + agent-written notes.","cited_cells":47,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/replit-agent--replit-com"},{"id":"replit-replit-md--docs-replit-com","name":"Replit replit.md","url":"https://docs.replit.com/replitai/replit-dot-md","section":"File-backed / editor paradigms","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Project-level Replit Agent context","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"file (replit.md)","llm_lock":"","github":"","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"manual","description":"replit.md at project root. Persists project overview, tech stack, architectural decisions, deployment process, external resources. Scoped to Agent conversations only.","cited_cells":30,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/replit-replit-md--docs-replit-com"},{"id":"researchrabbit--researchrabbit-ai","name":"ResearchRabbit","url":"https://www.researchrabbit.ai/","section":"Vertical / domain-specific AI memory","tier":2,"status":"acquired","last_verified_at":"2026-10-23","type":"Iterative paper-graph \"rabbit-hole\" checkpoints","license":"","latest_release":"","deployment":"Managed-only; acquired by Litmaps May 2025","pricing":"Free; RR+ $10/mo annual or $12.50/mo monthly (parity discounts in 100+ countries); Institution custom","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph","retrieval":"graph-traversal","persistence":"cross-session","update":"append-only","unit":"document","governance":"user-controllable","conflict":"append","description":"Acquired by Litmaps (announced 2025-05-08, terms undisclosed); product still running. Literature discovery + citation mapping. Oct 2025 revamp formalises checkpointed search paths as named \"rabbit holes\" — each checkpoint is a saved state of the citation graph the user can branch from. Navigational memory rather than stateless re-querying.","cited_cells":33,"third_party_cells":10,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/researchrabbit--researchrabbit-ai"},{"id":"retell-ai--retellai-com","name":"Retell AI","url":"https://www.retellai.com/","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"In-call + integration-based long-term","license":"","latest_release":"","deployment":"Managed-only (Retell-hosted; Enterprise dedicated-server add-on; no self-host option listed)","pricing":"Pay-as-you-go voice $0.07–0.31/min, chat from $0.002/msg, $10 free credits, 20 free concurrent calls (+$8/concurrent call/mo); Enterprise custom","mcp_support":"yes — MCP Node lets agents call tools on external MCP servers mid-call","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"multi-vendor","github":"","storage":"kv|vector","retrieval":"injection|similarity","persistence":"cross-session","update":"overwrite|extraction","unit":"episode|fact","governance":"inspectable","conflict":"none","description":"Voice agents. Short-term context within call native; long-term memory via webhook integration to CRM / external DB. Cross-agent memory sharing in multi-agent setups.","cited_cells":44,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/retell-ai--retellai-com"},{"id":"retro--arxiv-2112-04426","name":"RETRO","url":"https://arxiv.org/abs/2112.04426","section":"Retrieval-as-memory hybrids","tier":3,"status":"research-code","last_verified_at":"2026-10-23","type":"Trillion-token retrieval transformer","license":"not applicable — no official DeepMind code (unofficial lucidrains/RETRO-pytorch is Apache-2.0)","latest_release":"v0.3.9 (2023-10-30)","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"lucidrains/RETRO-pytorch (unofficial third-party impl.) — ~878 stars","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"read-only","unit":"chunk","governance":"opaque","conflict":"n/a","description":"DeepMind. Retrieval-Enhanced Transformer with chunked cross-attention over a 2T-token retrieval database. Frozen BERT retriever + differentiable encoder.","cited_cells":25,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/retro--arxiv-2112-04426"},{"id":"rivet-memory-node-ironclad--rivet-ironcladapp-com","name":"Rivet Memory Node (Ironclad)","url":"https://rivet.ironcladapp.com/","section":"Framework-embedded memory","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Visual-graph first-class memory node","license":"MIT","latest_release":"","deployment":"Both (OSS desktop app or self-hosted; remote executor for cloud/production deployment)","pricing":"OSS free (MIT)","mcp_support":"searched not found","a2a_support":"not supported","backend_storage":"","llm_lock":"any (BYO)","github":"ironclad/rivet — ~4.7k stars, last commit n/a","storage":"kv","retrieval":"exact-match","persistence":"session","update":"overwrite","unit":"kv-token","governance":"inspectable","conflict":"overwrite","description":"Open-source visual AI programming environment. Memory is a first-class graph node: stores text block, supports stable IDs (saving same ID overwrites enabling updates), injects stored text into downstream nodes when relevant. Memory is designed and debugged visually.","cited_cells":49,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/rivet-memory-node-ironclad--rivet-ironcladapp-com"},{"id":"roo-code-roorules--docs-roocode-com","name":"Roo Code .roorules","url":"https://roocodeinc.github.io/Roo-Code/features/custom-instructions/","section":"File-backed / editor paradigms","tier":2,"status":"archived","last_verified_at":"2026-10-29","type":"Per-mode rules + workspace overrides","license":"Apache-2.0","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"file (.roorules)","llm_lock":"","github":"RooCodeInc/Roo-Code — ~24.3k stars, archived (last push 2026-05-15)","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"manual","description":"Discontinued: Roo Code (VS Code agent) used .roorules / .roo/rules files as persistent project instructions; the extension shut down on 2026-05-15 and its repo is archived.","cited_cells":28,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/roo-code-roorules--docs-roocode-com"},{"id":"routerag--arxiv-2512-09487","name":"RouteRAG","url":"https://arxiv.org/abs/2512.09487","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"RL-routed text + graph + hybrid","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"","storage":"hybrid","retrieval":"similarity|graph-traversal|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"End-to-end RL-trained RAG. Token-level policy chooses when to retrieve and which mode (text / KG / hybrid via RRF). Two-stage Group-Relative Policy Optimisation (GRPO). No public repo at time of writing.","cited_cells":20,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/routerag--arxiv-2512-09487"},{"id":"salesforce-agentforce-einstein--salesforce-com","name":"Salesforce Agentforce / Einstein","url":"https://www.salesforce.com/artificial-intelligence/","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Customer 360 + Data Cloud as context layer","license":"","latest_release":"","deployment":"searched not found","pricing":"Flex Credits $500 per 100K credits; $2/conversation; user license $5/user/mo + credits; Agentforce add-ons $125–150/user/mo; Agentforce 1 Editions from $550/user/mo","mcp_support":"via official adapter — Salesforce MCP","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"custom (Salesforce platform)","llm_lock":"","github":"","storage":"relational|vector","retrieval":"similarity|injection","persistence":"long-term","update":"overwrite","unit":"fact|profile","governance":"auditable","conflict":"pii","description":"Agentforce agents grounded in Data Cloud — unified customer record across silos. Memory is the CRM record made available to AI; not an independent memory system. AI reads/writes through the existing data model.","cited_cells":36,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/salesforce-agentforce-einstein--salesforce-com"},{"id":"sanctuary-ai-carbon-phoenix-gen-7-8--sanctuary-ai","name":"Sanctuary AI Carbon (Phoenix Gen 7/8)","url":"https://www.sanctuary.ai/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-29","type":"Cognitive architecture with explicit task memory","license":"","latest_release":"","deployment":"Hardware deployment at customer sites","pricing":"Est. starting ~$65000 USD (configuration-dependent)","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"kv|vector","retrieval":"agentic|similarity","persistence":"long-term","update":"extraction","unit":"episode|skill|fact","governance":"opaque","conflict":"none","description":"Carbon control system originally built for the Phoenix humanoid; company now sells it as hardware-agnostic Physical AI for existing industrial robots plus in-development hydraulic hands, with the Phoenix humanoid no longer featured.","cited_cells":27,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/sanctuary-ai-carbon-phoenix-gen-7-8--sanctuary-ai"},{"id":"sandbar-stream--sandbar-com","name":"Sandbar Stream","url":"https://www.sandbar.com","section":"Voice-first / wearable AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Smart ring + voice-activated capture","license":"","latest_release":"","deployment":"Managed-only; still listed as preorder on site as of Oct 2026","pricing":"Ring preorder $249 (silver) / $299 (gold); Pro subscription $10/mo (free 3mo for preorders)","mcp_support":"yes — first-party Stream MCP server (api.sandbar.com/mcp, OAuth; Claude, Claude Code, ChatGPT, Codex); invite-only beta","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"","storage":"file","retrieval":"extraction-pull","persistence":"long-term","update":"append-only","unit":"episode","governance":"user-controllable","conflict":"none","description":"Smart ring with voice-activated capture (tap to record, not always-on). Personal knowledge graph with companies, people, meetings; pulls context from past meetings and shared interests. Ex-Meta engineering team.","cited_cells":32,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/sandbar-stream--sandbar-com"},{"id":"saner-ai--saner-ai","name":"Saner.ai","url":"https://www.saner.ai/","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"AI-auto-tagged + ADHD focus mode","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free (30 AI requests/mo, 100 notes); Starter $8/mo (30 AI req/day, 1K notes); Standard $16/mo (unlimited notes, 100 AI req/day); ~20% off yearly","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"extraction","unit":"document","governance":"user-controllable","conflict":"none","description":"Auto-tagged flat notes; semantic clusters (no folders). \"Skai\" personal AI surfaces related notes, semantic search, cross-tool integration (Drive, Slack, Calendar). ADHD-focused.","cited_cells":31,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/saner-ai--saner-ai"},{"id":"schoolai--schoolai-com","name":"SchoolAI","url":"https://schoolai.com/","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Teacher-configured Spaces with observable session logs","license":"","latest_release":"","deployment":"Web (K-12 classroom deployment)","pricing":"Free trial tier (5 Space launches/yr); Pro and Scale (district) tiers, prices on request","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"relational","retrieval":"injection","persistence":"cross-session","update":"append-only","unit":"episode","governance":"auditable","conflict":"append","description":"Teacher-created \"Spaces\" (structured AI workspaces) where Dot, the student-facing assistant, guides learners through agendas. Student interaction data — completion, errors, time-on-task — surfaces to teachers via dashboards. Memory is within-space session context; the teacher holds the persistent profile, not the AI itself.","cited_cells":36,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/schoolai--schoolai-com"},{"id":"scite-ai--scite-ai","name":"Scite.ai","url":"https://scite.ai/","section":"Vertical / domain-specific AI memory","tier":2,"status":"acquired","last_verified_at":"2026-10-22","type":"Intent-classified citation context index","license":"","latest_release":"","deployment":"Managed-only; acquired by Research Solutions Dec 2023","pricing":"Basic $12/mo (billed yearly, 7-day trial); Pro $40/mo (billed yearly); Enterprise custom","mcp_support":"yes — first-party Scite MCP for AI agents; MCP credits 250/mo (Basic), 2,500/mo (Pro)","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|relational","retrieval":"similarity|exact-match","persistence":"long-term","update":"extraction","unit":"document|fact","governance":"inspectable","conflict":"append","description":"Stores surrounding textual context of every citation plus a deep-learning classification (supports / disputes / mentions). Persisted across the database, not computed on demand. MCP integration (2025) exposes the memory layer to Claude and ChatGPT.","cited_cells":29,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/scite-ai--scite-ai"},{"id":"screenpipe--gh-mediar-ai-screenpipe","name":"Screenpipe","url":"https://github.com/mediar-ai/screenpipe","section":"Personal AI / PKM / lifelogging memory","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Product","license":"source-available (Screenpipe Commercial License)","latest_release":"last commit 2026-10-22","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/mediar-ai/screenpipe (22031★)","storage":"relational|file","retrieval":"exact-match|agentic","persistence":"long-term","update":"append-only","unit":"episode","governance":"user-controllable|inspectable","conflict":"append","description":"Local-first 24/7 screen and audio recorder that indexes what you see/hear and exposes it as context/memory to AI agents (MCP, API); open-source Rewind/Recall alternative. Y Combinator S26","cited_cells":15,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/screenpipe--gh-mediar-ai-screenpipe"},{"id":"scrintal--scrintal-com","name":"Scrintal","url":"https://scrintal.com/","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Whiteboard + linked cards (spatial graph)","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free; Pro $9/mo (annual) — from product claims row","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph","retrieval":"graph-traversal","persistence":"long-term","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"overwrite","description":"Infinite whiteboard + bidirectional-linked cards (spatial + networked hybrid). AI assistant for brainstorming, summarisation, rewriting selected cards.","cited_cells":30,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/scrintal--scrintal-com"},{"id":"second-me-mindverse--gh-mindverse-second-me","name":"Second Me (Mindverse)","url":"https://github.com/mindverse/Second-Me","section":"Personal AI / PKM / lifelogging memory","tier":3,"status":"stale","last_verified_at":"2026-10-28","type":"Product","license":"Apache-2.0","latest_release":"last commit 2025-09-19","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"mindverse/Second-Me — ~15.7k stars, last commit 2025-09","storage":"parametric|hybrid","retrieval":"parametric-recall|similarity","persistence":"parametric-permanent|lifelong","update":"parametric-edit","unit":"weight|profile|document","governance":"user-controllable","conflict":"training-time","description":"Open-source system for training a local 'AI self' from your own notes and memories using hierarchical memory modelling, deployable as a personal agent on a network of Second Mes. #12 in TeleAI Awesome-Agent-Memory","cited_cells":15,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/second-me-mindverse--gh-mindverse-second-me"},{"id":"self-rag--arxiv-2310-11511","name":"Self-RAG","url":"https://arxiv.org/abs/2310.11511","section":"Retrieval-as-memory hybrids","tier":4,"status":"research-code","last_verified_at":"2026-10-23","type":"Reflection-token self-critique","license":"MIT","latest_release":"no-release no-release","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"2.4k★ Python","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Trains an LLM end-to-end to emit reflection tokens that decide whether to retrieve, assess passage relevance, and critique output quality — all within a single model.","cited_cells":25,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/self-rag--arxiv-2310-11511"},{"id":"semantic-kernel-memory--learn-microsoft-com","name":"Semantic Kernel Memory","url":"https://learn.microsoft.com/en-us/semantic-kernel/frameworks/agent/agent-memory","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-29","type":"Memory plugin for SK agents","license":"MIT","latest_release":"python-1.45.0 (2026-10-06)","deployment":"Both (on-prem, containers, any cloud, or Azure AI Foundry managed)","pricing":"OSS (MIT); no direct cost for library; pay only for Azure/model consumption when using managed Azure infrastructure","mcp_support":"via official adapter — SK MCP plugin","a2a_support":"supported — SK A2A integration","backend_storage":"pluggable","llm_lock":"any (BYO via connectors)","github":"microsoft/semantic-kernel — ~28.4k stars, last commit 2026-10","storage":"vector","retrieval":"similarity|injection","persistence":"cross-session","update":"overwrite","unit":"document|chunk","governance":"inspectable","conflict":"overwrite","description":"Microsoft Semantic Kernel's memory plugin (Microsoft.SemanticKernel.Plugins.Memory): a plug-in approach to long-term memory in the SK agent framework (agent memory marked experimental). Semantic Kernel's README names Microsoft Agent Framework 1.0 as its enterprise-ready successor, with a migration guide.","cited_cells":61,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/semantic-kernel-memory--learn-microsoft-com"},{"id":"serena-oraios--gh-oraios-serena","name":"Serena (Oraios)","url":"https://github.com/oraios/serena","section":"Coding-agent memory","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"GPL-3.0-or-later (SolidLSP MIT)","latest_release":"last commit 2026-10-20","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/oraios/serena (30609★)","storage":"file","retrieval":"agentic|exact-match","persistence":"cross-session","update":"agent-controlled|overwrite","unit":"file|document","governance":"editable","conflict":"agent","description":"LSP-backed MCP coding toolkit ('IDE for your agent') with a built-in project memory system for sharing knowledge across sessions, users and projects alongside AGENTS.md.","cited_cells":16,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/serena-oraios--gh-oraios-serena"},{"id":"sg-nav--arxiv-2410-08189","name":"SG-Nav","url":"https://arxiv.org/abs/2410.08189","section":"Vertical / domain-specific AI memory","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Online 3D scene graph (incremental spatial memory)","license":"MIT","latest_release":"no releases","deployment":"Open-source (MIT)","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"bagh2178/SG-Nav — ~355 stars, last commit n/a","storage":"graph","retrieval":"graph-traversal","persistence":"session","update":"append-only","unit":"scene-graph","governance":"inspectable","conflict":"append","description":"Incrementally constructs hierarchical 3D scene graph during navigation — nodes for objects/groups/rooms, edges for spatial relations. Prunes stale nodes with linear complexity. Live graph passed as structured prompt to LLM that traverses with chain-of-thought.","cited_cells":35,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/sg-nav--arxiv-2410-08189"},{"id":"sierra--sierra-ai","name":"Sierra","url":"https://sierra.ai","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-23","type":"Unified customer graph (Agent Data Platform)","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"extraction","unit":"profile|episode","governance":"inspectable","conflict":"deterministic","description":"Bret Taylor + Clay Bavor. Agent Data Platform links unstructured conversation data to enterprise systems (billing, inventory, policies, transactions); persists across channels and time. Memory accumulates across every interaction, not just within a session.","cited_cells":59,"third_party_cells":8,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/sierra--sierra-ai"},{"id":"sigma-private-ai-browser--sigmabrowser-com","name":"Sigma (Private AI Browser)","url":"https://www.sigmabrowser.com/","section":"Browser-agent memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Fully-local LLM browser memory","license":"","latest_release":"","deployment":"Managed-only (browser app install)","pricing":"Free $0; Plus $10/mo ($8 annual); Pro $75/mo ($60 annual)","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"custom (local)","llm_lock":"","github":"","storage":"vector|file","retrieval":"similarity","persistence":"long-term","update":"append-only","unit":"episode|document","governance":"user-controllable","conflict":"none","description":"Chromium-based browser with AI agent (navigates pages, fills forms), chat-with-page, deep research and an in-browser local LLM (Eclipse) for private on-device processing; macOS, Windows, iOS, Android beta.","cited_cells":33,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/sigma-private-ai-browser--sigmabrowser-com"},{"id":"skild-brain--skild-ai","name":"Skild Brain","url":"https://www.skild.ai/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"In-context real-time memory + live data collection","license":"","latest_release":"","deployment":"Cloud inference + on-robot edge deployment","pricing":"No public pricing; $4k–$15k target hardware cost enabled by Skild Brain","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"attention|similarity","persistence":"session","update":"overwrite","unit":"episode","governance":"opaque","conflict":"overwrite","description":"Real-time simultaneous inference + data collection — dynamically updates in-context representation as the robot encounters new situations. Lets the system adapt to novel embodiments / unseen environments without explicit fine-tuning. Pre-trained on internet human video + physics sim.","cited_cells":29,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/skild-brain--skild-ai"},{"id":"smolagents-memory--huggingface-co","name":"smolagents Memory","url":"https://huggingface.co/docs/smolagents/tutorials/memory","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"In-process Python list (no persistence)","license":"Apache-2.0","latest_release":"v1.26.0 (2026-05-29)","deployment":"Both (OSS self-hosted; Hugging Face Inference Endpoints / Enterprise Hub managed)","pricing":"OSS free (Apache-2.0); Hugging Face Enterprise Hub starts at $20/user/mo for managed services","mcp_support":"via official adapter — smolagents.MCPClient","a2a_support":"searched not found","backend_storage":"in-memory","llm_lock":"any (BYO via HF Inference / OpenAI / local)","github":"huggingface/smolagents — ~28.8k stars","storage":"kv","retrieval":"injection","persistence":"session","update":"append-only","unit":"episode","governance":"inspectable","conflict":"append","description":"Hugging Face barebones agent framework. Memory lives as a Python list; no built-in persistence. Step callbacks let you mutate memory between steps. Persistence is left to the user — community proposals exist for \"memory bank\"-style additions.","cited_cells":56,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/smolagents-memory--huggingface-co"},{"id":"soul-machines--soulmachines-com","name":"Soul Machines","url":"https://www.soulmachines.com/","section":"Vertical / domain-specific AI memory","tier":2,"status":"archived","last_verified_at":"2026-10-29","type":"Patented Digital Brain + cross-session memory","license":"","latest_release":"","deployment":"Managed-only","pricing":"Enterprise only","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|kv","retrieval":"similarity|injection","persistence":"cross-session","update":"extraction","unit":"episode|fact","governance":"opaque","conflict":"none","description":"In receivership since 2026-02-05 (KPMG receivers Leon Bowker and Luke Norman, appointed under a security deed held for convertible-note holders; NZ Gazette notice 2026-ar623); website still up. NZ/US digital-humans company. Patented Digital Brain simulates emotional state and learns from interactions. Cross-session Memories layer for user- and persona-specific context persistence. Deployed by ANZ Bank (Jamie) and Mercedes-Benz.","cited_cells":30,"third_party_cells":5,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/soul-machines--soulmachines-com"},{"id":"sourcegraph-cody-cody-json--docs-sourcegraph-com","name":"Sourcegraph Cody — cody.json","url":"https://docs.sourcegraph.com/cody/custom-commands","section":"File-backed / editor paradigms","tier":2,"status":"deprecated","last_verified_at":"2026-10-23","type":"Custom command templates","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"file (cody.json)","llm_lock":"","github":"","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"manual","description":"Cody Free, Pro and Enterprise Starter were discontinued 2025-07-23 (Amp offered as replacement); only Cody Enterprise continues. .vscode/cody.json (workspace-scoped). JSON defining slash-commands with prompt text, context options ( currentFile , openTabs , terminal ), and mode ( ask / edit / insert ). Reusable prompt templates for project-specific workflows; checked in for team sharing.","cited_cells":31,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/sourcegraph-cody-cody-json--docs-sourcegraph-com"},{"id":"speak--speak-com","name":"Speak","url":"https://www.speak.com/","section":"Vertical / domain-specific AI memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Onboarding curriculum + per-session conversation","license":"","latest_release":"","deployment":"iOS + Android + Web + Enterprise embedded","pricing":"7-day free trial; Premium and Premium Plus subscriptions (price varies by region, shown in-app); Speak for Business","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"relational","retrieval":"injection","persistence":"cross-session","update":"overwrite|extraction","unit":"profile|episode","governance":"inspectable","conflict":"overwrite","description":"Speaking-first AI language app (Spanish, French, Korean, Japanese, Italian, Chinese). Uses the learner's mistakes and history to generate personalised review; Premium Plus adds 'Made for You' lessons from gaps and an AI Speak Tutor. Business edition with admin progress tracking.","cited_cells":32,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/speak--speak-com"},{"id":"speculative-rag--arxiv-2407-08223","name":"Speculative RAG","url":"https://arxiv.org/abs/2407.08223","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"Small-model draft, large-model verify","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"","storage":"vector","retrieval":"similarity|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Smaller distilled specialist generates multiple RAG drafts in parallel from distinct document subsets. Larger generalist verifies in a single pass. Subsets formed by content-similarity clustering to minimise redundancy.","cited_cells":23,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/speculative-rag--arxiv-2407-08223"},{"id":"spellbook-library--lawnext-com","name":"Spellbook Library","url":"https://www.spellbook.com/library","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-23","type":"Firm-precedent corpus memory","license":"","latest_release":"","deployment":"Managed SaaS inside Microsoft Word (plus Google Docs, iManage, SharePoint/OneDrive integrations)","pricing":"7-day free trial; paid pricing via sales (book a demo)","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"extraction","unit":"document","governance":"auditable","conflict":"none","description":"Transactional / contract drafting. Ingests the firm's precedent corpus (OneDrive, Dropbox or direct upload); indexes it so lawyers search past clauses in plain language inside Word, insert one, and have Spellbook adapt it to the style and context of the current deal. Goal: eliminate \"reads like ChatGPT\" output.","cited_cells":32,"third_party_cells":10,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/spellbook-library--lawnext-com"},{"id":"spirit-ai-character-engine--spiritai-com","name":"Spirit AI Character Engine","url":"https://www.spiritai.com/character-engine/","section":"Vertical / domain-specific AI memory","tier":2,"status":"archived","last_verified_at":"2026-10-29","type":"Author-defined character knowledge graph","license":"","latest_release":"","deployment":"searched not found","pricing":"searched not found","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph","retrieval":"graph-traversal","persistence":"cross-session","update":"overwrite","unit":"fact","governance":"user-controllable","conflict":"overwrite","description":"Defunct: Spirit AI Limited entered members' voluntary liquidation (liquidators appointed 2023-09-18). Authoring toolkit + SDK for narrative designers. Tracks what each character knows, doesn't know, has witnessed; explicit handling of NPC knowledge inconsistency across sessions. Used in AR/VR and Alexa-style voice game contexts.","cited_cells":25,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/spirit-ai-character-engine--spiritai-com"},{"id":"spring-ai-chatmemory--docs-spring-io","name":"Spring AI ChatMemory","url":"https://docs.spring.io/spring-ai/reference/api/chat-memory.html","section":"Framework-embedded memory","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Advisor-pattern Java-native memory","license":"Apache-2.0","latest_release":"v2.0.1 (2026-08-21)","deployment":"Both (on-prem, any cloud, or VMware Tanzu managed platform)","pricing":"OSS free (Apache-2.0); Tanzu Spring Enterprise subscription available for commercial support","mcp_support":"via official adapter — Spring AI MCP","a2a_support":"searched not found","backend_storage":"","llm_lock":"any (BYO via ChatModel)","github":"spring-projects/spring-ai — ~9.4k stars","storage":"kv","retrieval":"injection","persistence":"session","update":"append-only","unit":"document","governance":"inspectable","conflict":"append","description":"Wraps conversation memory behind an Advisor abstraction. MessageWindowChatMemory default 20-msg sliding window (preserves system messages); pluggable ChatMemoryRepository backends (in-memory, JDBC, Cassandra, Neo4j). Memory injected as cross-cutting concern at the ChatClient level.","cited_cells":52,"third_party_cells":10,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/spring-ai-chatmemory--docs-spring-io"},{"id":"stack-ai--stackai-com","name":"Stack AI","url":"https://www.stackai.com/","section":"Framework-embedded memory","tier":2,"status":"acquired","last_verified_at":"2026-10-29","type":"Enterprise-KB RAG (no agent memory store)","license":"","latest_release":"","deployment":"Multi-tenant cloud, VPC, or on-premise (Enterprise)","pricing":"Free $0 (500 runs/mo, 2 projects, 1 seat); Enterprise custom","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"multi-vendor","github":"","storage":"vector","retrieval":"similarity","persistence":"long-term","update":"overwrite","unit":"document","governance":"auditable","conflict":"n/a","description":"Acquired by Asana (completed 2026-05-28); continues as its own product and brand. Connects enterprise knowledge sources (SharePoint, Confluence, Notion, Drive, DBs) as memory layer with access controls, versioning, citations. Agents query at runtime; no separate session/cross-session store.","cited_cells":50,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/stack-ai--stackai-com"},{"id":"strands-agents-memory-aws--docs-aws-amazon-com","name":"Strands Agents Memory (AWS)","url":"https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/strands-sdk-memory.html","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Mem0 + AgentCore-backed","license":"Apache-2.0","latest_release":"","deployment":"Both (OSS self-hosted; AWS Bedrock AgentCore managed cloud)","pricing":"Strands SDK OSS free; AgentCore Memory: short-term $1.00/GB ingested from 2026-10-06 (was $0.25/1k events); long-term $0.75/1k records-mo, $0.50/1k retrievals","mcp_support":"native (first-party) — Strands MCP support","a2a_support":"supported — Strands A2A server","backend_storage":"pluggable (AgentCore, Mem0)","llm_lock":"multi-vendor (Bedrock/Anthropic/Gemini/OpenAI)","github":"strands-agents/sdk-python — ~8.9k stars, last commit n/a","storage":"vector|graph","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction","unit":"fact|episode","governance":"opaque","conflict":"llm-arbitrate","description":"AWS code-first agent framework. Multi-LLM (Bedrock, Anthropic, Gemini, OpenAI). Memory via Mem0 integration (ElastiCache for Valkey + Neptune Analytics) or Bedrock AgentCore Memory. Local default uses FAISS; OpenSearch for production.","cited_cells":52,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/strands-agents-memory-aws--docs-aws-amazon-com"},{"id":"structrag--arxiv-2410-08815","name":"StructRAG","url":"https://arxiv.org/abs/2410.08815","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"Task-adaptive structurisation","license":"no-license","latest_release":"n/a","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"Li-Z-Q/StructRAG — ~171 stars","storage":"hybrid","retrieval":"hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Identifies optimal structure (table, catalog, algorithm, graph) per task; converts source documents into that structure at inference time and reasons over the structured form. Motivated by cognitive theories of human knowledge conversion.","cited_cells":24,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/structrag--arxiv-2410-08815"},{"id":"suki-ai--suki-ai","name":"Suki AI","url":"https://www.suki.ai/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Pre-visit summary + prior-note synthesis","license":"","latest_release":"","deployment":"Managed-only","pricing":"SaaS","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|relational","retrieval":"similarity|injection","persistence":"long-term","update":"extraction|consolidation","unit":"episode|document","governance":"auditable","conflict":"clinician","description":"Doctor-assist ambient documentation. Pulls forward prior documentation into a pre-visit synthesis rather than waiting for the encounter. Patient summaries surface relevant insight from history before visit begins.","cited_cells":55,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/suki-ai--suki-ai"},{"id":"superlocalmemory--superlocalmemory-com","name":"SuperLocalMemory","url":"https://www.superlocalmemory.com/","section":"Dedicated memory layers","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Local-geometric / privacy-first / no LLM scoring","license":"AGPL-3.0-or-later (commercial licence available)","latest_release":"v4.2.10 (2026-10-17)","deployment":"Local-only (no Docker; no graph DB; no API keys; CPU-only; Mode A zero-LLM)","pricing":"free / open source (AGPL-3.0); commercial licence terms unpriced","mcp_support":"yes — stdio server (slm mcp) and local HTTP /mcp endpoint, configurable tool profiles","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"qualixar/superlocalmemory — ~227 stars, last commit 2026-10","storage":"vector|graph","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction","unit":"fact|episode","governance":"user-controllable","conflict":"none","description":"Cloud-free agent memory using information-geometric retrieval (Fisher-Rao distance + BM25 + entity graph + temporal fusion via 4-channel RRF) instead of LLM calls for scoring and lifecycle. Mode A is the only system to cross 70% LoCoMo with zero API dependency.","cited_cells":38,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/superlocalmemory--superlocalmemory-com"},{"id":"supermemory--supermemory-ai","name":"Supermemory","url":"https://supermemory.ai/","section":"Dedicated memory layers","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Memory graph + extraction + RAG","license":"MIT","latest_release":"server-v0.0.9 (2026-10-12); Python SDK v5.0.0 (2026-10-06)","deployment":"Both","pricing":"OSS","mcp_support":"native (first-party) — claude-supermemory plugin + MCP","a2a_support":"not documented publicly","backend_storage":"custom","llm_lock":"","github":"supermemoryai/supermemory — ~31.3k stars, last commit 2026-10","storage":"graph|vector","retrieval":"similarity|graph-traversal|extraction-pull","persistence":"long-term","update":"extraction","unit":"document|fact","governance":"opaque","conflict":"llm-arbitrate","description":"Memory engine with API, app, browser extension, and MCP server. Extracts facts, tracks updates, resolves contradictions, auto-forgets expired info. Plugins for Claude Code, OpenCode, OpenClaw, Hermes.","cited_cells":73,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/supermemory--supermemory-ai"},{"id":"supernotes-ai--supernotes-app","name":"Supernotes AI","url":"https://supernotes.app/","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Notecards + tags + AI tagging","license":"","latest_release":"","deployment":"Managed-only","pricing":"Starter free (100 cards); Unlimited £6/mo or £72/yr (25% off); 4-year plan 50% off","mcp_support":"no official server — team building OAuth 2.1 for a hosted MCP (Jun 2026); API usable by LLMs today","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"","storage":"relational|vector","retrieval":"similarity","persistence":"long-term","update":"overwrite|extraction","unit":"document","governance":"user-controllable","conflict":"overwrite","description":"Flat notecards + tags + collections; bidirectional references. AI tag recommendations, grammar suggestions, Vision (notecard editing assistant), Collection Creator.","cited_cells":31,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/supernotes-ai--supernotes-app"},{"id":"superpowers-episodic-memory-plugin--gh-obra-episodic-memory","name":"Superpowers episodic-memory plugin","url":"https://github.com/obra/episodic-memory","section":"Claude Code memory mechanisms","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"MCP semantic search over conversations","license":"MIT","latest_release":"v1.4.2 (2026-05-21)","deployment":"local — runs as Claude Code plugin via superpowers marketplace; no cloud component","pricing":"free / open source (MIT); no paid tier","mcp_support":"native (first-party) — provides MCP tools to Claude Code","a2a_support":"not supported","backend_storage":"searched not found","llm_lock":"","github":"obra/episodic-memory — ~473 stars, last commit 2026-05","storage":"vector","retrieval":"similarity","persistence":"cross-session","update":"append-only","unit":"episode","governance":"inspectable","conflict":"n/a","description":"Claude Code plugin that indexes prior conversations at session end and exposes search/show tools via MCP. Distributed via obra/superpowers-marketplace ; install with /plugin install episodic-memory@superpowers-marketplace .","cited_cells":44,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/superpowers-episodic-memory-plugin--gh-obra-episodic-memory"},{"id":"sweep-ai--sweep-dev","name":"Sweep AI","url":"https://sweep.dev","section":"Coding-agent memory","tier":2,"status":"deprecated","last_verified_at":"2026-10-29","type":"GitHub-thread-as-implicit-memory","license":"proprietary (Sweep Enterprise Edition License; some parts MIT)","latest_release":"","deployment":"Both (GitHub App managed + JetBrains plugin local)","pricing":"Free (community); Pro and Team tiers; LLM cost + 5% markup model for agent usage","mcp_support":"searched not found — Sweep is a GitHub-app-shaped tool; MCP not surfaced","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"sweepai/sweep — ~7.7k stars (legacy GitHub-app code)","storage":"file","retrieval":"extraction-pull","persistence":"session","update":"read-only","unit":"episode","governance":"inspectable","conflict":"none","description":"Superseded: the sweepai/sweep README says the team now builds an AI coding assistant for JetBrains; the GitHub-app product described here is no longer the focus (repo not archived). No dedicated persistent memory layer. Operates as a GitHub App — reads issue + codebase fresh on each trigger. Some context implicit in the PR/issue threads it created previously. Successor product: Sweep: AI Autocomplete and Coding Agent JetBrains plugin (Marketplace id 26860; latest version 1.29.3, 2026-02-05). As of Oct 2026 sweep.dev does not resolve to a website.","cited_cells":38,"third_party_cells":8,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/sweep-ai--sweep-dev"},{"id":"synthesis-tutor--synthesis-com","name":"Synthesis Tutor","url":"https://www.synthesis.com/tutor","section":"Vertical / domain-specific AI memory","tier":3,"status":"active","last_verified_at":"2026-10-22","type":"Real-time micro-assessment adaptation","license":"","latest_release":"","deployment":"iPad + Android tablet + Web/desktop (incl. Chromebook)","pricing":"7-day free trial; family ~$29/mo or $119/yr; individual $45/mo or ~$300/yr; lifetime $999–$1,499 (page figures inconsistent)","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"","storage":"kv","retrieval":"injection","persistence":"session","update":"overwrite","unit":"profile","governance":"inspectable","conflict":"n/a","description":"Embeds micro-assessments into every lesson to track understanding in real time, adjusting difficulty and teaching approach within session. Combines AI personalization with human-curated content. No publicly disclosed mechanism for a durable cross-session learner profile beyond progress tracking within its math curriculum scope (ages 5–11).","cited_cells":29,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/synthesis-tutor--synthesis-com"},{"id":"tabnine--docs-tabnine-com","name":"Tabnine","url":"https://docs.tabnine.com/main/administering-tabnine/managing-your-team/context-engine","section":"Coding-agent memory","tier":2,"status":"acquired","last_verified_at":"2026-10-29","type":"Org-level Context Engine + Tabnine Agents","license":"","latest_release":"","deployment":"Both","pricing":"No free tier; Code Assistant $39/user/mo, Agentic Platform $59/user/mo (annual); Headless Agents add-on $1.2k–5k/mo","mcp_support":"yes — Tabnine Agent in IDE (VS Code/JetBrains) and Tabnine CLI connect to stdio/HTTP/SSE MCP servers via mcpServers in settings.json or `tabnine mcp add`","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"graph|vector","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"extraction","unit":"document|fact","governance":"opaque","conflict":"none","description":"Enterprise Context Engine builds a multi-dimensional knowledge graph from repos + CI/CD + PRs + tickets + docs at the org level. Individual completions stay session-scoped. Tabnine Agents (Nov 2025) adds team-shared agent context. Acquired by Tricentis (2026-07-30); Tricentis is integrating the Enterprise Context Engine into its Agentic Quality Engineering Platform, and tabnine.com now redirects to tricentis.com.","cited_cells":41,"third_party_cells":8,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/tabnine--docs-tabnine-com"},{"id":"tana--tana-inc","name":"Tana (now Tana Outliner)","url":"https://outliner.tana.inc/","section":"Personal AI / PKM / lifelogging memory","tier":2,"status":"renamed","last_verified_at":"2026-10-29","type":"Supertag-based PKM with AI","license":"","latest_release":"","deployment":"Managed-only","pricing":"Tana Outliner: Free (500 AI credits); Plus $8/mo; Pro $14/mo (yearly saves 22%). Meetings product: Free; Pro ~$20–30; Business from $35/user/mo","mcp_support":"yes — Outliner desktop runs local MCP server (localhost:8262/mcp, free); hosted remote MCP beta at app.tana.inc/mcp on paid plans","a2a_support":"searched not found","backend_storage":"custom","llm_lock":"","github":"","storage":"relational|graph","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"overwrite","unit":"fact|document","governance":"user-controllable","conflict":"overwrite","description":"Tana split in Mar 2026: the supertag-based outliner/PKM is now 'Tana Outliner' at outliner.tana.inc, while tana.inc is a separate agentic meeting platform sharing a knowledge graph. Outliner keeps Tana AI on nodes (chat with notes, summaries, automations).","cited_cells":46,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/tana--tana-inc"},{"id":"tc-rag--arxiv-2408-09199","name":"TC-RAG","url":"https://arxiv.org/abs/2408.09199","section":"Retrieval-as-memory hybrids","tier":3,"status":"research-code","last_verified_at":"2026-10-22","type":"Turing-complete state-variable RAG","license":"no-license","latest_release":"no-release no-release","deployment":"searched not found","pricing":"searched not found","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"73★ Python","storage":"hybrid","retrieval":"similarity|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Frames RAG as Turing-complete with explicit monitored state variables, enabling adaptive retrieval halting and convergence control. Medical-LLM case study.","cited_cells":25,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/tc-rag--arxiv-2408-09199"},{"id":"tencent-ima-copilot--ima-qq-com","name":"Tencent ima.copilot","url":"https://ima.qq.com","section":"Personal AI / PKM / lifelogging memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Four-module self-evolving personal memory","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|graph","retrieval":"similarity|graph-traversal","persistence":"long-term","update":"extraction","unit":"document|fact|episode","governance":"opaque","conflict":"none","description":"Tencent's knowledge-management productivity AI, powered by Hunyuan. Copilot mode adds a four-module persistent memory architecture: Soul (agent persona), User (profile), Memory (long-term declarative), Agent (learned skills/task experience). Memory persists across devices and sessions; context awareness reads active web pages and files automatically.","cited_cells":32,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/tencent-ima-copilot--ima-qq-com"},{"id":"tencentdb-agent-memory--gh-tencent-tencentdb-agent-memory","name":"TencentDB Agent Memory","url":"https://github.com/Tencent/TencentDB-Agent-Memory","section":"Dedicated memory layers","tier":3,"status":"active","last_verified_at":"2026-10-23","type":"Framework","license":"MIT","latest_release":"last commit 2026-10-21","deployment":"searched not found","pricing":"searched not found","mcp_support":"searched not found","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"searched not found","github":"https://github.com/Tencent/TencentDB-Agent-Memory (29800★)","storage":"vector|graph","retrieval":"hybrid-rerank|graph-traversal|agentic","persistence":"long-term","update":"extraction|consolidation","unit":"fact|skill|document","governance":"user-controllable|editable","conflict":"document-version","description":"Tencent's team-level memory hub that turns conversations, docs and code into four reusable memory assets: chat memory, skills, an LLM-wiki and a code graph. big-tech vendor","cited_cells":15,"third_party_cells":1,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/tencentdb-agent-memory--gh-tencent-tencentdb-agent-memory"},{"id":"tesla-fsd-v13-occupancy-4d-world--thinkautonomous-ai","name":"Tesla FSD V13 — Occupancy + 4D World (now FSD v14)","url":"https://www.thinkautonomous.ai/blog/occupancy-networks/","section":"Vertical / domain-specific AI memory","tier":1,"status":"renamed","last_verified_at":"2026-10-23","type":"3D occupancy network + temporal multi-modal fusion","license":"","latest_release":"FSD (Supervised) v14.3.6 for HW4 + v14 Lite for HW3 (software 2026.20.6.11, 2026-07-21) — latest found; proprietary","deployment":"On-device in Tesla vehicles; v14.x targets HW4/AI4","pricing":"FSD (Supervised) $99/month subscription; subscription-only path on Tesla support page; requires FSD computer 3.0+","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|kv-cache","retrieval":"attention","persistence":"session","update":"overwrite","unit":"episode|scene-graph","governance":"opaque","conflict":"overwrite","description":"Single end-to-end neural network ingesting multi-camera video + navigation + ego-motion, maintaining an evolving 3D occupancy volume as internal world state — spatial memory persisting across frames, tracking objects that temporarily leave the camera FOV. Row originally described V13; current line is FSD (Supervised) v14 — v14.3 (2026-04-07) rewrote the AI compiler and runtime with MLIR (~20% faster reaction time) and upgraded the RL stage and vision encoder; a v14 Lite build brings V14 to HW3 cars (2026-07).","cited_cells":29,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/tesla-fsd-v13-occupancy-4d-world--thinkautonomous-ai"},{"id":"trace--ycombinator-com","name":"Trace","url":"https://www.ycombinator.com/companies/trace-so","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Context-graph / KG memory","license":"","latest_release":"","deployment":"Cloud (SaaS integrating Slack/Jira/Notion; no self-hosted option found)","pricing":"not public — beta with waitlist / contact","mcp_support":"not documented publicly","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"","storage":"graph","retrieval":"graph-traversal","persistence":"long-term","update":"extraction","unit":"fact|episode","governance":"inspectable","conflict":"none","description":"YC S25 + $3M seed (Feb 2026). Workflow orchestration mapping corporate environments as a structured context graph so agents always have institutional knowledge for tasks. CTO frames differentiator as \"context engineering.\"","cited_cells":31,"third_party_cells":10,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/trace--ycombinator-com"},{"id":"vapi-voice-agents--vapi-ai","name":"Vapi (voice agents)","url":"https://vapi.ai/","section":"Framework-embedded memory","tier":2,"status":"active","last_verified_at":"2026-10-28","type":"Structured workflow state during call","license":"","latest_release":"","deployment":"Both","pricing":"Pay-as-you-go $0.05/min Vapi hosting with $5 free credit; Core $29/mo; Pro 10% of the hosting fee ($999/mo minimum); Premier custom","mcp_support":"via official adapter — Vapi MCP tools","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"multi-vendor","github":"","storage":"kv","retrieval":"injection","persistence":"session","update":"overwrite","unit":"fact","governance":"inspectable","conflict":"overwrite","description":"Voice agents. Externalises memory from LLM into structured call state — variables rather than model recall — reducing hallucinations. Squad mode passes context across multi-agent assistant hand-offs in same call.","cited_cells":46,"third_party_cells":4,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/vapi-voice-agents--vapi-ai"},{"id":"vektor-memory--dev-to","name":"VEKTOR Memory","url":"https://dev.to/vektor_memory_43f51a32376/vektor-openai-agents-sdk-production-memory-in-three-lines-59p6","section":"Dedicated memory layers","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Local-SQLite + AUDN-loop + TypeScript-native","license":"proprietary (commercial licence)","latest_release":"v1.7.8 Preview (2026-07-16)","deployment":"Local-only (Node.js; Transformers.js via WebAssembly; SQLite; no cloud infrastructure)","pricing":"Solo $9/mo (3 licences); Team $35/mo (5); Studio $59/mo (10); Enterprise $99/mo (25)","mcp_support":"yes — bundled MCP server (npx vektor mcp) with 50+ tools; Claude Code/Desktop integration","a2a_support":"not documented publicly","backend_storage":"searched not found","llm_lock":"","github":"Vektor-Memory/Vektor-memory — ~5 stars, latest changelog 2026-07","storage":"relational|vector","retrieval":"similarity","persistence":"cross-session","update":"extraction","unit":"fact|episode","governance":"inspectable","conflict":"llm-arbitrate","description":"Node.js/TypeScript-native memory layer. SQLite + local Transformers.js embeddings (~80MB model via WebAssembly). AUDN loop (Add/Update/Delete/None) resolves contradictions before write rather than at retrieval — read-after-write consistency.","cited_cells":33,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/vektor-memory--dev-to"},{"id":"vertex-ai-memory-bank--docs-cloud-google-com","name":"Vertex AI Memory Bank (now Agent Platform Memory Bank)","url":"https://docs.cloud.google.com/gemini-enterprise-agent-platform/scale/memory-bank","section":"Platform-provider memory","tier":2,"status":"renamed","last_verified_at":"2026-10-29","type":"Managed Vertex memory service","license":"","latest_release":"","deployment":"Managed-only","pricing":"Pay-per-use","mcp_support":"via official adapter (ADK)","a2a_support":"supported — A2A native","backend_storage":"custom (Google-managed)","llm_lock":"","github":"","storage":"vector","retrieval":"similarity","persistence":"cross-session","update":"extraction","unit":"fact","governance":"opaque","conflict":"none","description":"Google Cloud managed memory service, now 'Agent Platform Memory Bank' in Gemini Enterprise Agent Platform. LLM-based memory extraction/consolidation from sessions (background, multimodal, custom topics), per-identity isolation, similarity search, TTL, revisions; used from ADK via VertexAiMemoryBankService.","cited_cells":71,"third_party_cells":7,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/vertex-ai-memory-bank--docs-cloud-google-com"},{"id":"voiceflow-memory--docs-voiceflow-com","name":"Voiceflow Memory","url":"https://www.voiceflow.com/docs/documentation/build/behaviour","section":"Framework-embedded memory","tier":1,"status":"active","last_verified_at":"2026-10-23","type":"Sliding-window buffer (vf_memory)","license":"","latest_release":"","deployment":"Managed-only","pricing":"Free trial + usage-based billing for agencies/partners; business pricing on request","mcp_support":"yes — MCP tool lets agents call tools on remote MCP servers; Voiceflow also hosts an MCP server (mcp.voiceflow.com) for managing projects from AI clients","a2a_support":"searched not found","backend_storage":"searched not found","llm_lock":"multi-vendor (OpenAI/Anthropic/Gemini/Cohere)","github":"","storage":"kv","retrieval":"injection","persistence":"session","update":"overwrite","unit":"fact","governance":"user-controllable","conflict":"last-write-wins","description":"Agent platform. Keeps a configurable window of recent turns (default 50, range 10–100) in the {vf_memory} variable; older turns are summarized rather than dropped. Chat projects can persist history indefinitely or clear it on tab close/refresh.","cited_cells":50,"third_party_cells":3,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/voiceflow-memory--docs-voiceflow-com"},{"id":"wayve-gaia-2-gaia-3--wayve-ai","name":"Wayve GAIA-2 / GAIA-3","url":"https://wayve.ai/thinking/gaia-3/","section":"Vertical / domain-specific AI memory","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Multi-view latent diffusion world model","license":"","latest_release":"","deployment":"Research/internal tool — not sold separately","pricing":"searched not found","mcp_support":"no MCP support advertised — vertical product, no MCP server / client integration documented","a2a_support":"no A2A protocol support advertised — vertical product, no A2A integration documented","backend_storage":"searched not found","llm_lock":"","github":"","storage":"vector|kv-cache","retrieval":"attention|cache-lookup","persistence":"session","update":"append-only","unit":"episode|kv-token","governance":"opaque","conflict":"append","description":"All surround-camera views encoded through video tokeniser to continuous latent; past latent sequences serve as explicit temporal context via space-time factorised transformer. GAIA-3 (15B params, 2× GAIA-2 size, 10× training data) adds safety-critical scenario generation + embodiment transfer.","cited_cells":51,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/wayve-gaia-2-gaia-3--wayve-ai"},{"id":"weknow-rag--arxiv-2408-07611","name":"WeKnow-RAG","url":"https://arxiv.org/abs/2408.07611","section":"Retrieval-as-memory hybrids","tier":4,"status":"active","last_verified_at":"2026-10-22","type":"Sparse-dense web + KG self-assessment","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"","llm_lock":"","github":"","storage":"hybrid","retrieval":"similarity|graph-traversal|hybrid-rerank","persistence":"long-term","update":"read-only","unit":"chunk","governance":"inspectable","conflict":"n/a","description":"Domain KGs + multi-stage web-page retrieval (sparse + dense) + LLM self-assessment of trustworthiness before finalising answer.","cited_cells":20,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/weknow-rag--arxiv-2408-07611"},{"id":"windsurf-rules-memories--docs-windsurf-com","name":"Windsurf Rules & Memories (now Devin Desktop)","url":"https://docs.devin.ai/desktop/cascade/memories","section":"File-backed / editor paradigms","tier":1,"status":"acquired","last_verified_at":"2026-10-29","type":"Codebase semantic + auto-memory","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"file (.windsurfrules / global memories)","llm_lock":"","github":"","storage":"file|vector","retrieval":"injection|similarity","persistence":"cross-session","update":"overwrite|extraction","unit":"document|fact","governance":"user-controllable","conflict":"llm-arbitrate","description":"Now Devin Desktop (ex-Windsurf). Rules: global_rules.md (6K chars) plus workspace rules in .devin/rules/ (fallback .windsurf/rules/, 12K chars each) and AGENTS.md via the same engine. Cascade auto-memories stored in ~/.codeium/windsurf/memories/ are legacy-agent only, with a migration path to skills.","cited_cells":46,"third_party_cells":2,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/windsurf-rules-memories--docs-windsurf-com"},{"id":"zed-rules--zed-dev","name":"Zed .rules","url":"https://zed.dev/docs/ai/rules","section":"File-backed / editor paradigms","tier":2,"status":"active","last_verified_at":"2026-10-22","type":"Reads competitors as fallback","license":"","latest_release":"","deployment":"","pricing":"","mcp_support":"","a2a_support":"","backend_storage":"file (.rules)","llm_lock":"","github":"","storage":"file","retrieval":"injection","persistence":"cross-session","update":"overwrite","unit":"document","governance":"user-controllable","conflict":"manual","description":"Zed reads one project instruction file (first match of .rules, .cursorrules, AGENTS.md, CLAUDE.md, GEMINI.md and similar) into agent context. Since v1.4.0 the on-demand Rules Library has been replaced by Skills, and old rules migrate to global skills or AGENTS.md.","cited_cells":29,"third_party_cells":0,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/zed-rules--zed-dev"},{"id":"zep-governance-posture--getzep-com","name":"Zep — governance posture","url":"https://www.getzep.com/product/agent-memory/","section":"Memory governance, privacy & safety","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Bi-temporal KG = structural provenance","license":"","latest_release":"","deployment":"SaaS cloud; self-hosted (Graphiti OSS); Enterprise: BYOC VPC","pricing":"Free 10k credits/mo; Flex ~$125/mo (50k credits); Flex Plus ~$375/mo (200k credits); Enterprise custom (third-party tracker, prices inconsistent — confirm on getzep.com/pricing)","mcp_support":"native (first-party) — Graphiti MCP","a2a_support":"no data — searched getzep.com/product/agent-memory, help.getzep.com, docs.getzep.com; A2A protocol not advertised","backend_storage":"Postgres + Neo4j","llm_lock":"","github":"","storage":"graph","retrieval":"graph-traversal","persistence":"long-term","update":"append-only","unit":"fact|episode","governance":"auditable","conflict":"bi-temporal","description":"Bi-temporal knowledge graph preserves full lineage of every stored fact — every memory write traceable and auditable by graph schema, not bolted-on logging. SOC 2 Type II; HIPAA BAA available; BYOK + VPC-isolated deployment for data residency.","cited_cells":46,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/zep-governance-posture--getzep-com"},{"id":"zep-graphiti--getzep-com","name":"Zep & Graphiti","url":"https://www.getzep.com/","section":"Dedicated memory layers","tier":1,"status":"active","last_verified_at":"2026-10-22","type":"Bi-temporal knowledge graph","license":"Apache-2.0","latest_release":"v0.30.0 (2026-09-01) (graphiti-core; MCP server mcp-v1.1.0 same day)","deployment":"Both — Zep Cloud managed; Graphiti OSS self-hosted (Neo4j/FalkorDB)","pricing":"Free 10k credits/mo; Flex $125/mo ($104 annual) 50k credits; Flex Plus $375/mo ($312 annual) 200k credits; Enterprise custom","mcp_support":"native (first-party) — Graphiti MCP server","a2a_support":"not documented publicly","backend_storage":"Postgres + Neo4j (Graphiti)","llm_lock":"","github":"getzep/graphiti — ~32k stars, last commit 2026-09","storage":"graph","retrieval":"graph-traversal|similarity","persistence":"long-term","update":"append-only","unit":"episode|fact","governance":"auditable","conflict":"bi-temporal","description":"Bi-temporal knowledge graph (event time + ingestion time). Strong on chronological reasoning and contradiction tracking. Graphiti is the open-source core.","cited_cells":81,"third_party_cells":6,"page":"https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/zep-graphiti--getzep-com"}]}