| Guide | What it covers |
|---|---|
| ../README.md | Product overview, quickstart, install-by-surface, integration snippets, and examples. |
| usage.md | Full API reference: storage, learning, recall, MCP, integrations, multi-agent memory, and bi-temporal recall. |
| multi-agent-governance.md | The governed-shared-memory surface: access control, supersession, provenance, trust, and audit — mapped to code and to the 2026 literature. |
| memory-commands.md | Cross-surface command map for add, search, list, get, delete, learn, and structured correction. |
| integrations.md | Fast routing by ecosystem surface: CrewAI, LangGraph, LangChain, LlamaIndex, AutoGen, OpenAI Agents SDK, PydanticAI, MCP, HTTP, and TypeScript. |
| deployment.md | Install, storage backends, MCP server, HTTP sidecar, browser inspector, observability, security, and scaling. |
| air-gapped-deployment.md | Zero-network guarantee (with a test that enforces it), the complete network surface, offline install, and offline audit export for regulated / air-gapped environments. |
| troubleshooting.md | First-run, install, MCP, npm bridge, and Pages troubleshooting. |
| codespaces-quickstart.md | Install-free first-run flow for GitHub Codespaces. |
| ../examples/README.md | Runnable example index across the core loop, integrations, MCP, HTTP, browser, and TypeScript paths. |
| ../skills/README.md | Assistant-facing skill surface for coding tools that support repo skills. |
| Guide | What it covers |
|---|---|
| positioning.md | One-liner, ICP, pain solved, why now, differentiation, and non-goals. |
| comparison.md | Buyer-facing "when to choose ai-knot vs Mem0 / Graphiti / Letta / LangMem". |
| faq.md | Short answers for objections, evaluation questions, and benchmark skepticism. |
| benchmarks.md | Reproducible benchmark methodology, deterministic retrieval suite, LoCoMo, and LongMemEval. |
| production-readiness.md | Current readiness across storage, retrieval, integrations, observability, CI, and release. |
| Guide | What it covers |
|---|---|
| whitepaper.md | Research-style long-form argument for treating memory as a knowledge layer instead of a log. |
| developer-article.md | Practical developer article for adding deterministic memory to an agent quickly. |
| site/index.html | GitHub Pages landing page. |
| site/whitepaper.html | GitHub Pages version of the whitepaper. |
| site/developer-article.html | GitHub Pages version of the developer article. |
| ../scripts/render_site_articles.py | Rebuilds the Pages article HTML from the Markdown sources. |
| ../scripts/render_whitepaper_pdf.py | Renders the whitepaper Markdown into a PDF artifact when needed. |
| Guide | What it covers |
|---|---|
| crewai-case-study.md | CrewAI memory object, scoped agents, and proof flow. |
| langgraph-case-study.md | LangGraph / LangChain tool-style memory surfaces and when to use them. |
| llamaindex-case-study.md | LlamaIndex memory=... seam and deterministic long-term memory underneath it. |
| autogen-case-study.md | AutoGen memory adapter path and usage shape. |
| openai-agents-case-study.md | OpenAI Agents SDK RunConfig seam for recalled memory. |
| pydanticai-case-study.md | Runtime instructions=... memory injection for PydanticAI. |
| vercel-ai-sdk-case-study.md | TypeScript / Vercel AI SDK system-prompt memory path. |
| openclaw-case-study.md | OpenClaw MCP setup and persistent-memory path. |
| claude-mcp-case-study.md | Claude Desktop / Claude Code MCP setup and proof flow. |
| http-sidecar-case-study.md | HTTP-first memory loop, sidecar path, and browser-inspector flow. |
| Guide | What it covers |
|---|---|
| RELEASE.md | Release workflow, version sync, registry publishing, and pre-release checks. |
| ../server.json | MCP Registry manifest for the PyPI-backed ai-knot server. |