# AI Agent Infrastructure Landscape

> An analysis tool for the AI agent stack: it helps you choose the components of an agent or multi-agent system and check a stack you already run. Underneath is a citation-backed catalog of 1002 systems: agent frameworks, models, memory layers, vector and graph stores, gateways, guardrails, identity and MCP tooling, and observability. Every cell links to its source and carries a verifiability tier. 2542 typed edges connect the systems. Data is CC-BY-4.0, as of 2026-10-31.

## Answer a builder's question

- [Build a whole agent stack](https://mrpeppersdev.github.io/agent-infrastructure-landscape/build): product type + requirements → one cited pick per layer (framework, model, gateway, memory, vector store, tools, sandbox, evals, guardrails…), why not the runners-up, exit paths, documented integrations, a risk score, OWASP LLM Top 10 coverage and a stack.yml. MCP: `plan_stack`; CLI: `landscape stack`.
- [Choose a memory layer](https://mrpeppersdev.github.io/agent-infrastructure-landscape/choose), [an agent framework](https://mrpeppersdev.github.io/agent-infrastructure-landscape/choose/framework) or [a vector / graph database](https://mrpeppersdev.github.io/agent-infrastructure-landscape/choose/data): a few questions → three cited picks.
- [Check a stack](https://mrpeppersdev.github.io/agent-infrastructure-landscape/check): paste package.json / requirements.txt / pyproject.toml / MCP config / stack.yml → archived, acquired or stale dependencies and replacements.
- [Alternatives](https://mrpeppersdev.github.io/agent-infrastructure-landscape/alternatives): live alternatives to any memory layer, framework or data store.
- [Ask the catalog](https://mrpeppersdev.github.io/agent-infrastructure-landscape/search): plain-language search with requirement checking.
- [Memory & vector cost calculator](https://mrpeppersdev.github.io/agent-infrastructure-landscape/costs): monthly cost of managed memory layers and vector stores from published prices.
- [Benchmarks](https://mrpeppersdev.github.io/agent-infrastructure-landscape/benchmarks): reproducible memory benchmark runs vs vendor-reported scores.
- [Compatibility](https://mrpeppersdev.github.io/agent-infrastructure-landscape/compatibility): do these systems work together? Integration evidence for every pair, with sources.
- [Compare](https://mrpeppersdev.github.io/agent-infrastructure-landscape/compare): head-to-head pages; [compare up to five](https://mrpeppersdev.github.io/agent-infrastructure-landscape/compare/many).
- [Findings](https://mrpeppersdev.github.io/agent-infrastructure-landscape/findings): headline findings from the catalog.

## Catalogs (one audience at a time; the home table is every system)

- [Agent Memory](https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/memory): Shipping memory layers, memory features and memory products for AI agents. Table: https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/memory/browse
- [Agent Frameworks & Harnesses](https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/agents): Frameworks, harnesses, IDEs and runtimes that agents are built and run in. Table: https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/agents/browse
- [Retrieval & Data Infrastructure](https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/retrieval): Vector stores, knowledge graphs, embeddings and search that memory is built on. Table: https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/retrieval/browse
- [Models, Serving & Ops](https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/models): Foundation models, inference, training, evaluation and observability around agents. Table: https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/models/browse
- [Memory Research & Benchmarks](https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/research): Papers, benchmarks and ideas behind agent memory, before they ship as products. Table: https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/research/browse
- [Coding Agents](https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/coding): IDEs, coding harnesses, the memory they keep and the sandboxes they run in. Table: https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/coding/browse
- [Security, Governance & Identity](https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/security): Guardrails, memory governance, agent identity and MCP gateways. Table: https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/security/browse
- [Evals & Observability](https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/evals): Evaluation platforms, tracing, memory monitoring and the benchmarks behind them. Table: https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/evals/browse
- [Voice, Computer-use & Embodied Agents](https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/embodied): Voice agent platforms, wearable memory, computer-use agents and robotics stacks. Table: https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/embodied/browse
- [Assistant & Personal Memory](https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/assistants): How ChatGPT, Claude, Gemini and personal-knowledge tools remember you. Table: https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/assistants/browse
- [Vertical & Domain Agents](https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/vertical): Agents and memory built for one industry: legal, health, support, sales, science, gaming. Table: https://mrpeppersdev.github.io/agent-infrastructure-landscape/catalogs/vertical/browse

## Machine-readable data (static JSON, no auth)

- [System index](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/index.json): every system with id, family, section, lifecycle status and a link to its JSON.
- One system: `https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/systems/<id>.json` — every cell with value, citation, status and tier; taxonomy; edges; change history; lifecycle.
- [Agent Memory](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/memory.json) (311 systems; [CSV](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/memory.csv))
- [Memory Research & Benchmarks](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/memory-research.json) (264 systems; [CSV](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/memory-research.csv))
- [Agent Frameworks & Harnesses](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/agents.json) (235 systems; [CSV](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/agents.csv))
- [Retrieval & Data Infrastructure](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/data.json) (58 systems; [CSV](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/data.csv))
- [Models, Serving & Ops](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/models-ops.json) (134 systems; [CSV](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/models-ops.csv))
- [Coding Agents](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/coding.json) (96 systems; [CSV](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/coding.csv))
- [Security, Governance & Identity](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/security.json) (42 systems; [CSV](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/security.csv))
- [Evals & Observability](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/evals.json) (72 systems; [CSV](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/evals.csv))
- [Voice, Computer-use & Embodied Agents](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/embodied.json) (53 systems; [CSV](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/embodied.csv))
- [Assistant & Personal Memory](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/assistants.json) (47 systems; [CSV](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/assistants.csv))
- [Vertical & Domain Agents](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/vertical.json) (150 systems; [CSV](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/vertical.csv))
- [Full catalog, flattened](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/all.json) ([CSV](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/catalogs/all.csv)); canonical source: [landscape.json](https://github.com/MrPeppersDev/agent-infrastructure-landscape/raw/main/data/landscape.json), [edges](https://github.com/MrPeppersDev/agent-infrastructure-landscape/raw/main/data/landscape.edges.json).
- [Stack-checker index](https://mrpeppersdev.github.io/agent-infrastructure-landscape/api/stack-index.json): npm / PyPI / Docker package and GitHub repo → system and lifecycle status.

## Change feeds

- [What changed](https://mrpeppersdev.github.io/agent-infrastructure-landscape/changes): acquisitions, shutdowns, renames, price and release changes, by week. [Atom feed](https://mrpeppersdev.github.io/agent-infrastructure-landscape/changes.xml); one system: `https://mrpeppersdev.github.io/agent-infrastructure-landscape/feeds/<id>.xml`.
- [Recently verified (Atom)](https://mrpeppersdev.github.io/agent-infrastructure-landscape/feed.xml).

## Tools

- [MCP server](https://github.com/MrPeppersDev/agent-infrastructure-landscape/tree/main/mcp): ask_catalog, choose_memory_layer, choose_framework_or_store, check_stack, check_compatibility, find_alternatives, system_history, estimate_memory_cost, estimate_vector_store_cost, search_records, compare, recent_changes and more.
- [CLI](https://github.com/MrPeppersDev/agent-infrastructure-landscape/tree/main/cli): `landscape check package.json --fail-on-attention` for CI.
- [GitHub Action](https://github.com/MrPeppersDev/agent-infrastructure-landscape/tree/main/action): `uses: MrPeppersDev/agent-infrastructure-landscape@main` flags archived, acquired or stale AI dependencies on pull requests.
- [Hosted MCP endpoint](https://github.com/MrPeppersDev/agent-infrastructure-landscape/tree/main/worker): remote, read-only MCP server (Streamable HTTP).

## Optional

- [About: how to read a record](https://mrpeppersdev.github.io/agent-infrastructure-landscape/about)
- [Data & API docs](https://mrpeppersdev.github.io/agent-infrastructure-landscape/data)
- [Submit a system](https://mrpeppersdev.github.io/agent-infrastructure-landscape/submit) · [Correct a record](https://github.com/MrPeppersDev/agent-infrastructure-landscape/issues/new?template=correction.yml)
- [Sitemap](https://mrpeppersdev.github.io/agent-infrastructure-landscape/sitemap.xml)