Local-first MCP server that turns any public GitHub, GitLab, or Bitbucket repo into fresh, version-pinned documentation for AI coding assistants.
ctxpeek is easy to set up with strong trust signals. Check agent compatibility and use-case fit before adding it to your workflow.
Repository setup guidance
gh repo view tamimbinhakim/ctxpeek --webOpen the official README and confirm the supported install method.
Add the server entry to your MCP client config.
Restart your agent and verify that the server tools appear.
ctxpeek is a tool that lets AI coding assistants read documentation directly from public code repositories. It fetches the exact files from a specific version of a repo, caches them locally, and returns them as markdown. No accounts, no cloud services, no third-party instructions needed.
ctxpeek (pronounced Context Peek) is a local stdio MCP server that provides AI coding assistants with up-to-date, version-pinned documentation directly from public GitHub, GitLab, and Bitbucket repositories. It resolves fuzzy package names (e.g., 'drizzle orm' → drizzle-team/drizzle-orm) via package registries, then lists the repository's documentation tree and fetches specific files at a given commit, branch, or tag. The tool is local-first: all caching, listing, and fetching happen on the user's machine, with no telemetry or third-party services. It returns markdown instead of JSON, reducing token usage by ~75%. It supports monorepo subpaths and conditional HTTP requests to minimize rate limit usage. Built for agentic clients, ctxpeek enables models to navigate documentation structure directly rather than relying on vector search or chunk dumps. The project is open-source under Apache 2.0 and welcomes contributions.
Strong trust signals; still review the README and permissions before production use.
Last commit was about 52 days ago.
0 GitHub stars indicate community interest.
10 open issues signal maintenance load.
Apache-2.0 license detected.
Ask an AI coding assistant to show you the routing documentation for Next.js v15.0.0 directly from the official repo.
Get the latest API docs for a library you're using, pinned to a specific commit or tag, without leaving your editor.
Resolve a fuzzy package name like 'drizzle orm' to the correct repository and fetch its documentation tree.
Integrate with Claude Code, Cursor, or Windsurf to provide on-demand, versioned documentation for any open-source project.
Use as a privacy-focused documentation tool that never sends your queries to a third-party server.
The tool fetches data from public repositories; ensure you have permission to access the repos you query.
While local-first, the initial fetch for uncached repos requires network access to the hosting platform.
No authentication is required for public repos, but using a GitHub PAT may increase rate limits.
0
Stars
0
Forks
10
Issues
Apache-2.0
License
Optimize markdown documentation for LLMs and RAG systems, reducing token consumption by 67-95%.
Codebase intelligence layer for AI coding agents: code health, git analytics, dead code detection, and architectural decisions via MCP.
Turn OpenAPI, MCP, Doxygen, godoc, rustdoc, and Markdown into static documentation sites you own.
Local-first MCP server that turns any public GitHub, GitLab, or Bitbucket repo into fresh, version-pinned documentation for AI coding assistants.
AI coding assistant users who need accurate, versioned documentation from public repos., Developers using Claude Code, Cursor, Windsurf, or other MCP-compatible agents., Open-source maintainers who want to provide fresh docs to AI tools without extra infrastructure.
The setup section provides repository-level starting guidance, not a guarantee of an independently verified installation. Check the official README and release notes.
The tool fetches data from public repositories; ensure you have permission to access the repos you query. While local-first, the initial fetch for uncached repos requires network access to the hosting platform.
Similar or complementary options to evaluate include llm-docs-builder, repowise, sourcey.
Jun 7, 2026
3 security/trust notes recorded.
Setup difficulty is 2/5.