The 50 most-recently verified systems in the v6 catalog. https://mrpeppersdev.github.io/agent-infrastructure-landscape/feed.xml 2026-06-30T00:00:00Z Mr. Peppers https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/atlas--arxiv-2208-03299 2026-06-30T00:00:00Z

Few-shot retrieval-augmented LM · Retrieval-as-memory hybrids · Meta AI. Jointly fine-tunes retriever + LM for few-shot tasks.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/compressive-transformer--arxiv-1911-05507 2026-06-30T00:00:00Z

Recent full-res + compressed older · Recent method papers — theorized, no distinct product · Maintains recent states in full resolution while compressing older memories with learned compression functions. DeepMind.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/emat--arxiv-2210-16773 2026-06-30T00:00:00Z

Sub-ms KV memory for QA pairs · Recent method papers — theorized, no distinct product · Efficient Memory-Augmented Transformer. Encodes millions of QA pairs into key-value memory; MIPS search returns hits in sub-millisecond latency. EMNLP 2022.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/flare--arxiv-2305-06983 2026-06-30T00:00:00Z

Confidence-triggered forward retrieval · Retrieval-as-memory hybrids · 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.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/generative-agents--arxiv-2304-03442 2026-06-30T00:00:00Z

Interactive simulacra of human behavior · Recent method papers — theorized, no distinct product · Park et al. — landmark agent-simulation paper. Reflection + memory stream + retrieval enable believable agent behavior.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/iter-retgen--arxiv-2305-15294 2026-06-30T00:00:00Z

Output-conditioned iterative retrieval · Retrieval-as-memory hybrids · 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.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/memorybank--arxiv-2305-10250 2026-06-30T00:00:00Z

Long-term memory enhancement · Recent method papers — theorized, no distinct product · Enhances LLMs with long-term memory. Early influential paper.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/neural-episodic-control--arxiv-1703-01988 2026-06-30T00:00:00Z

Differentiable episodic dictionary · Recent method papers — theorized, no distinct product · Pritzel et al. Agent stores past state-action-value tuples in a differentiable dictionary; recall via approximate nearest neighbour. ICML 2017.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/retro--arxiv-2112-04426 2026-06-30T00:00:00Z

Trillion-token retrieval transformer · Retrieval-as-memory hybrids · DeepMind. Retrieval-Enhanced Transformer with chunked cross-attention over a 2T-token retrieval database. Frozen BERT retriever + differentiable encoder.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/transformer-xl--arxiv-1901-02860 2026-06-30T00:00:00Z

Segment-level recurrence + cached states · Recent method papers — theorized, no distinct product · Extends context through segment-level recurrence + caching of hidden states from prior segments. Foundational long-context architecture.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/bytebot--bytebot-ai 2026-05-15T00:00:00Z

OSS desktop-control agent runtime (containerised Linux desktop) · Computer-use & desktop agents · Open-source desktop-agent runtime that runs an LLM-controlled Linux desktop inside Docker — uses xdotool / X11 capture for control. Pitched as the 'self-hosted Operator' alternative; agent sees a real desktop screen, not just a browser.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/ewc-elastic-weight-consolidation--arxiv-1612-00796 2026-05-15T00:00:00Z

Quadratic constraint per weight · Recent method papers — theorized, no distinct product · Kirkpatrick et al. Soft quadratic constraint pulls each weight back toward old values, scaled by importance for prior tasks. PNAS 2017.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/gpt-engineer--gh-antonosika-gpt-engineer 2026-05-15T00:00:00Z

Spec-to-codebase agent · Agent frameworks (no first-party memory layer) · One of the original 'agent writes a whole codebase from a prompt' demos — community-maintained; spawned the commercial Lovable (gptengineer.app).

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/i-jepa--arxiv-2301-08243 2026-05-15T00:00:00Z

Joint-embedding predictive arch (image) · Recent method papers — theorized, no distinct product · Meta. Self-supervised image learning by predicting target-block representations from a context block in latent space — no hand-crafted augmentations or pixel reconstruction. Foundation of LeCun's hierarchical-JEPA programme. CVPR 2023.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/mem0-mcp-official--gh-mem0ai-mem0-mcp 2026-05-15T00:00:00Z

Hybrid vector + graph via cloud API · Claude Code memory mechanisms · 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.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/text-generation-inference-tgi--gh-huggingface-text-generation-inference 2026-05-15T00:00:00Z

HF inference server · Inference platforms & gateways · Hugging Face's production inference server for LLMs — continuous batching, FlashAttention, quantization (bitsandbytes, GPTQ).

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/0-5-physical-intelligence--pi-website 2026-05-14T00:00:00Z

Multi-Scale Embodied Memory (short + long episodic) · Vertical / domain-specific AI memory · 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.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/01-ai-yi-family--01-ai 2026-05-14T00:00:00Z

Open-weights frontier-tier Chinese model family (Yi-Lightning / Yi-34B) · Foundation models (substrate reference) · Kai-Fu Lee's 01.AI (founded 2023, Beijing). Yi family open-weights — Yi-34B (2023), Yi-Large (proprietary 2024), Yi-Lightning (2024-10, frontier-tier, matched GPT-4o on LMSYS chatbot arena). Apache 2.0 (Yi-34B and base sizes). Important second Chinese open-weights option alongside DeepSeek + Qwen.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/11x-ai--11x-ai 2026-05-14T00:00:00Z

AI sales agents (Alice, Mike, Jordan) · Use-case-specific agent harnesses · Founded 2022 by Hasan Sukkar; AI sales agents Alice (outbound SDR), Mike (voice SDR), Jordan (RevOps). $50M Series B Oct-2024 (Benchmark + Andreessen Horowitz; $350M val). Among the most-discussed AI-employee startups of 2024.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/1x-technologies--1x-tech 2026-05-14T00:00:00Z

Humanoid robot maker (Neo / Eve) — OpenAI portfolio · Robotics foundation models & agent stacks · Norway / US humanoid robot company — Eve (wheeled bimanual) + Neo (legged humanoid, consumer-targeted, late 2025 / 2026 launch). Raised $100M Series B (Jan-2024, EQT + OpenAI Startup Fund); $40M Series A (Mar-2023, OpenAI lead). Distinguishing feature: consumer-home positioning rather than industrial.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/6sense--6sense-com 2026-05-14T00:00:00Z

ABM intent + AI agent platform · Use-case-specific agent harnesses · Founded 2013; ABM/intent platform with Revenue AI agents. $200M Series E Jan-2022 ($5.2B val; Blue Owl). 6sense Revenue AI agents launched 2024 — autonomous account research + email drafting.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/a-mem--gh-wujiangxu-a-mem 2026-05-14T00:00:00Z

Atomic-note / Zettelkasten-style · Research / specialised systems · Treats memories as atomic linkable notes — explicit nod to Zettelkasten knowledge management. Dynamic linking; retroactive memory revision.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/abridge--abridge-com 2026-05-14T00:00:00Z

Grounded-transcript provenance · Vertical / domain-specific AI memory · 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.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/accelerate-hugging-face--gh-huggingface-accelerate 2026-05-14T00:00:00Z

Distributed training abstraction · Training infrastructure · Hugging Face's abstraction over distributed-training backends (DDP, FSDP, DeepSpeed, Megatron). Minimal code change to scale a PyTorch script.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/acon--arxiv-2510-00615 2026-05-14T00:00:00Z

Context-compression optimisation · Recent method papers — theorized, no distinct product · Optimises context compression for long-horizon LLM agents.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/activeloop-deep-lake--activeloop-ai 2026-05-14T00:00:00Z

Multimodal vector + serverless Postgres · Vector-database infrastructure · Deep Memory feature optimises embedding space per use-case (+22% retrieval accuracy). Deep Lake PG unifies serverless Postgres (agent short-term state) + billion-scale vector search (long-term memory).

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/acuvity-now-proofpoint--acuvity-ai 2026-05-14T00:00:00Z

MCP / Shadow-AI runtime enforcement · Memory governance, privacy & safety · 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.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/adaptive-rag--gh-starsuzi-adaptive-rag 2026-05-14T00:00:00Z

Query-complexity routing · Retrieval-as-memory hybrids · Smaller classifier LM predicts query complexity, then routes to no-retrieval / single-step / iterative retrieval as appropriate. NAACL 2024.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/adept-act--adept-ai 2026-05-14T00:00:00Z

Computer-use / workflow-automation agent (Adept now part of Amazon AGI 2024) · Multi-agent orchestration platforms · Founded 2022 by ex-OpenAI / Google researchers (David Luan, Kelsey Schroeder). Built ACT-1 / ACT-2 multimodal action transformers for computer-use. **Acquired by Amazon June-2024 (acqui-hire)** — co-founders + key team joined Amazon AGI; Adept the company continues with Zach Brock as remaining executive. Important historical entry — ACT models inspired Anthropic Computer Use + OpenAI Operator.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agency-swarm--gh-vrsen-agency-swarm 2026-05-14T00:00:00Z

OpenAI-Assistants-based multi-agent · Agent frameworks (no first-party memory layer) · Agent framework built on OpenAI Assistants API — agencies (roles + comm flow); replaced after Assistants v2 changes.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agent-kb--arxiv-2507-06229 2026-05-14T00:00:00Z

Cross-domain experience KB · Recent method papers — theorized, no distinct product · Leverages cross-domain experience for agentic problem solving.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agent-reconfigure-reconfigurable-agent--arxiv-2410-21465 2026-05-14T00:00:00Z

Skill-library agent memory · Recent method papers — theorized, no distinct product · Agent that maintains a skill library and reconfigures its toolset per task — procedural memory of successful sub-task solutions.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agent-s--arxiv-2410-08164 2026-05-14T00:00:00Z

Open agentic computer-use framework · Recent method papers — theorized, no distinct product · Open agentic framework that uses computers like a human.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agent-workflow-memory--openreview-net 2026-05-14T00:00:00Z

Workflow as memory · Recent method papers — theorized, no distinct product · Workflow-based memory framework component.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agent2agent-protocol-a2a--google-github-io 2026-05-14T00:00:00Z

Open agent-to-agent protocol · Agent frameworks (no first-party memory layer) · Google-led open protocol for agent-to-agent collaboration (April 2025) — agent cards, RPCs, server-sent events. Partner companies: Atlassian, MongoDB, Salesforce, ServiceNow, etc.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agentevolver--arxiv-2511-10395 2026-05-14T00:00:00Z

Efficient self-evolving agent system · Recent method papers — theorized, no distinct product · Towards an efficient self-evolving agent system.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agentfold--arxiv-2510-24699 2026-05-14T00:00:00Z

Proactive context management · Recent method papers — theorized, no distinct product · Long-horizon web agents with proactive context management.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agentic-memory--arxiv-2601-01885 2026-05-14T00:00:00Z

Unified short + long-term management · Recent method papers — theorized, no distinct product · Learning unified long-term and short-term memory management.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agentic-plan-caching-apc--neurips-cc 2026-05-14T00:00:00Z

Test-time plan-template memory · Recent method papers — theorized, no distinct product · Extracts, stores, adapts, and reuses structured plan templates from planning stages of agent applications. NeurIPS 2025 poster.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agentops--agentops-ai 2026-05-14T00:00:00Z

Operational metrics for agent memory · Memory observability & monitoring · 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.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agents-md--agents-md 2026-05-14T00:00:00Z

Cross-tool open standard · File-backed / editor paradigms · AGENTS.md at repo root (and ~/.codex/AGENTS.md for global); hierarchical concatenation from root to cwd. Persists build/test commands, conventions, architecture overview, security constraints, git workflow.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agility-robotics--agilityrobotics-com 2026-05-14T00:00:00Z

Bipedal warehouse robot maker (Digit) — first US humanoid commercial deployment · Robotics foundation models & agent stacks · Oregon-based humanoid maker; Digit is in production at GXO Logistics warehouses since Sep-2024 — claimed first humanoid in commercial service in the US. Spinout of Oregon State University. $400M+ raised (Amazon Industrial Innovation Fund participant).

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agixt-adaptive-memory--gh-josh-xt-agixt 2026-05-14T00:00:00Z

Adaptive provider-spanning + plugin storage · Framework-embedded memory · 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.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agno-phidata-memory--docs-phidata-com 2026-05-14T00:00:00Z

AgentStorage + AgentMemory + DBs · Framework-embedded memory · 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.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/agntcy-internet-of-agents--agntcy-org 2026-05-14T00:00:00Z

Cisco-LangChain agent-mesh protocols · Agent frameworks (no first-party memory layer) · Cisco + LangChain-led 'Internet of Agents' collective — open standards for discovery, messaging, observability of agent networks.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/ai-persona--arxiv-2412-13103 2026-05-14T00:00:00Z

Life-long LLM personalisation · Recent method papers — theorized, no distinct product · Life-long personalisation of LLMs.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/ai-singapore-sea-lion--sea-lion-ai 2026-05-14T00:00:00Z

SE-Asian cross-lingual semantic-retrieval memory · Dedicated memory layers · 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.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/aider--aider-chat 2026-05-14T00:00:00Z

Per-repo chat history file + YAML config · Coding-agent memory · 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.

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/aider-conventions-md-aider-conf-yml--aider-chat 2026-05-14T00:00:00Z

Multi-file convention pattern · File-backed / editor paradigms · CONVENTIONS.md (any name, loaded via --read or read: in config); .aider.conf.yml (home / git-root / cwd, last wins); .aiderignore (gitignore-syntax).

https://mrpeppersdev.github.io/agent-infrastructure-landscape/systems/aider-harness--aider-chat 2026-05-14T00:00:00Z

Terminal-native AI pair programmer (OSS, Python) · Agent IDEs & coding harnesses · OSS terminal coding agent by Paul Gauthier — one of the original AI-pair-programmer CLIs (2023). Architect/Editor modes; repo-map for context; uses any LiteLLM-compatible model. Memory via CONVENTIONS.md and .aider.conf.yml — author-edited markdown. Distinct from the separate Aider memory-row in "Coding-agent memory"; this row characterises the harness itself.