Get This Tool
OpenWiki
Pricing
- Model
- Free
Summary
AI coding agents hallucinate context about your codebase because there is no context to give them — AGENTS.md is empty, CLAUDE.md doesn't exist, and the agent is guessing at architecture from filenames. OpenWiki is a CLI that reads your codebase and writes that missing documentation, then keeps it from going stale.
OpenWiki runs as a CLI tool — `npm install -g openwiki`, run `--init` to configure your model and API key, and it generates documentation written for agents to consume rather than humans to read. The included GitHub Actions workflow opens a daily pull request with documentation updates, so the gap between your code and your AGENTS.md doesn't compound silently over time. The tool is built by langchain-ai and targets repositories already using LangChain or similar agent frameworks. Where it breaks: the page describes no fine-grained control over which files or modules get documented, and teams with large monorepos or sensitive internal APIs will need to audit what the LLM is reading before that daily PR becomes a liability.
Bottom line: The right call for a greenfield LangChain project where AGENTS.md is currently blank — a harder sell for a monorepo where you need scoped, audited documentation generation and the tool gives you a single global sweep.
Community Performance Report Card
No community ratings yet. Be the first to rate this tool!
Community Benchmarks Community
Sign in to submit a benchmarkNo community benchmarks yet. Be the first to share a real-world data point.
Pros
Sign in to edit- MIT-licensed and self-hostable, so the LLM API calls stay in your infrastructure and never route through a vendor's servers — which matters when your codebase contains IP you cannot send to a third-party pipeline.
- Generates output in AGENTS.md and CLAUDE.md conventions, so agent tools that rely on those files get populated context immediately rather than operating on empty or stale files that cause hallucinated architectural assumptions.
- Daily GitHub Actions PR keeps documentation synchronized after code changes, so the agent context your team ships tomorrow reflects the refactor that merged yesterday — without anyone remembering to update the docs manually.
- Provider-agnostic model configuration, so you point it at whatever LLM your organization has already approved rather than being locked into a specific API contract.
Cons
Sign in to edit- The page describes no scoping or exclusion configuration — the tool reads your codebase as a whole. Teams with large monorepos or modules containing credentials, internal API details, or proprietary logic have no documented way to exclude directories from the LLM sweep, which means a manual audit layer sits between `openwiki --init` and trusting the output in production.
- The daily PR cadence is fixed by the provided GitHub Actions workflow. Teams that need documentation updates triggered by specific events — a merge to main, a version tag, a changed module — must rewrite the workflow themselves, adding maintenance overhead that scales with how far their requirements drift from the default.
- There is no output format customization described in the page. Teams whose agents expect structured frontmatter, section schemas, or domain-specific documentation templates find that OpenWiki's output is shaped by the LLM's defaults, not their standards — at which point teams with strong documentation conventions switch to a scripted prompt pipeline they control directly.
Community Reviews
Sign in to write a reviewNo reviews yet. Be the first to share your experience.
About
- Platforms
- npm / Node.js
- API Available
- No
- Self-Hosted
- Yes
- Last Updated
- 2026-07-06T13:23:31.481Z
Best For
Who it's for
- Repositories using LangChain or similar agent frameworks
- Teams that rely on AI coding agents for maintenance
- Projects needing automated daily doc updates via GitHub Actions
What it does well
- Generating initial codebase documentation for agent consumption
- Keeping documentation synchronized after code changes
- Appending agent prompts to AGENTS.md or CLAUDE.md
Integrations
Discussion Community
Sign in to commentNo discussion yet. Sign in to start the conversation.
Compare OpenWiki
Spotted incorrect or missing data? Join our community of contributors.
Sign Up to ContributeCommunity Notes & Tips Community
Sign in to contributeBe the first to contribute. General notes, observations, gotchas, and tips from people who use this tool day-to-day.
Frequently Asked Questions
- Is OpenWiki free?
- Yes — OpenWiki is fully free to use. There is no paid tier.
- Is OpenWiki open source?
- Yes. OpenWiki is open source.
- Can I self-host OpenWiki?
- Yes. OpenWiki supports self-hosting on your own infrastructure.
- What platforms does OpenWiki support?
- OpenWiki is available on: npm / Node.js.
Hours Saved & ROI Stories Community
Sign in to contributeBe the first to contribute. Concrete time/cost savings, with context. e.g. "Cut my code review backlog from 4h to 45m per week."
Curated lists that include this category
OpenWiki is a command-line tool that generates and maintains codebase documentation targeted specifically at AI coding agents. The core workflow is a three-step sequence: initialize via `openwiki –init`, configure your LLM provider and API key, then run the generator. Output lands in files like AGENTS.md and CLAUDE.md — the convention files that tools like Claude Code and similar agents read to orient themselves before touching your code.
The differentiating feature is the GitHub Actions integration. The included `openwiki-update.yml` workflow opens a pull request once a day with updated documentation, which means documentation drift — the gap between what the code does and what AGENTS.md says it does — gets surfaced on a fixed cadence rather than discovered when an agent does something wrong. That daily PR is both the value proposition and the review surface: a human signs off before stale documentation ships.
OpenWiki fits tightly into a specific workflow: teams using agent-assisted development who have already committed to the AGENTS.md and CLAUDE.md convention and want that documentation generated rather than hand-written. It does not fit teams who need fine-grained control over documentation scope, exclusion lists, or output structure. The vendor states no hosted API and no paid tiers — the MIT license means you can fork and adapt, but the docs describe no configuration surface for telling the tool what to skip. A monorepo with proprietary internals requires that audit step before trusting a daily automated sweep.