# OpenWiki

> Updated 2026-09-16 · type: tool · category: productivity-knowledge · status: active · rev 1

OpenWiki is a CLI that writes and maintains agent-readable docs of your codebase, so coding agents stop re-learning your repo from scratch on every task.

- Open source: yes (MIT)
- Self-hostable: yes
- Pricing model: free
- Best for: Teams whose codebase has outgrown what an agent can hold in one context window — where every task starts with the agent re-reading the repo — and who want a maintained doc layer it can load instead, ideally runnable against a local model.
- Last verified: 2026-09-16

- **Canonical:** https://gtmstacker.com/registry/tool/openwiki/
- **Source:** [github · langchain-ai/openwiki](https://github.com/langchain-ai/openwiki)
- **Tags:** productivity-knowledge, developer-tooling, agent-memory, documentation, self-hostable
- **Repository:** https://github.com/langchain-ai/openwiki

## Is OpenWiki open source?

Yes, OpenWiki is open source under the MIT license.

## How much does OpenWiki cost?

OpenWiki is free to use.

## Can I self-host OpenWiki?

Yes, OpenWiki can be self-hosted (the source is available under the MIT license).

## Alternatives & related

- [Open Code Review](https://gtmstacker.com/registry/tool/open-code-review/)


---

MIT-licensed TypeScript CLI from LangChain that writes and maintains agent-readable documentation of your codebase, giving a coding agent persistent context instead of making it re-derive the repo on every task. It stores state locally in ~/.openwiki and operates on local repositories, and it can run through local model backends (Ollama, LM Studio, any OpenAI-compatible endpoint) as well as hosted providers, so a paid key is optional rather than required.

## Provenance

- MIT, TypeScript, ~16.6k stars, ~404 commits, local-state model and local-backend support independently WebFetch-verified on the repo 2026-09-16 (github.com/langchain-ai/openwiki, published under the langchain-ai org).
- Surfaced via the 2026-09-16 viral-posts brief (a post framing it as codebase memory for coding agents); catalogued on the verified repo facts.
- Curated from the GTM Stacker signal registry (2026-09-16 pass); license independently verified 2026-09-16.

## Why it matters for a GTM stack

The quiet tax on agent-built GTM tooling is context: past a certain repo size, every task begins with the agent re-reading the codebase, burning tokens and still missing things. OpenWiki's bet is to keep a maintained, agent-readable doc layer the agent can load instead — persistent memory of what the repo is and how it fits together. It is MIT, runs against local models, and comes from LangChain, which is a credible source for this exact problem. The catch is the failure mode of any generated-docs tool: stale docs can steer an agent wrong just as confidently as good ones help, so this is a keep-it-current layer, not a set-and-forget export. For a team maintaining more agent-written code than any one context can hold, that trade is usually worth making.