#7676th of 86 in MCP servers for developers
repobrain
Per-module AI agents that answer codebase questions with file paths and line numbers
- Stars
- 1.4k
- License
- MIT
- Last commit
- Oct 2026
- Last release
- Jul 2026
- Language
- Python
1,058 on 3 Apr 20261,356 on 30 Oct 2026
From GitHub's daily star counts.
Checked on GitHub.
Overview
RepoBrain runs `rb-refresh` to have model agents read each code module and write knowledge docs into a `.repobrain/` folder in the repo. `rb-ask` then routes a question to the relevant module agent and answers with source references. It works as a CLI, as Claude Code and Codex CLI plugins, or through an optional `rb-mcp` server for other agents and IDEs.
Who it is for: Developers who want coding agents to answer grounded questions about large repos
Strengths
- Knowledge lives in `.repobrain/` inside the repo, so several IDEs and agents share it
- Refresh is incremental after commits and resumes interrupted runs, keeping the last good generation
- Native plugins for Claude Code and Codex CLI, plus CLI/JSON output and optional MCP for other hosts
- Answers cite file paths and line numbers; README links a benchmark on four repos
Weaknesses
- Needs a configured model backend (OpenAI-compatible API or a logged-in headless CLI such as Trae)
- First build can take several minutes or longer on large repos and consumes model tokens
- Requires a Git repo with at least one commit and a clean worktree before refresh
- Answers lag behind code until `rb-refresh` is rerun; `rb-ask` only warns when stale
What it needs
- no GPU
- Docker + Compose
- Needs Git, OpenAI-compatible API or headless CLI backend (e.g. Trae)
- Models: OpenAI-compatible API, Headless CLI backends (Trae)
Quick answers
Is repobrain open source?
Yes. Its code is on GitHub under the MIT license.
Can you run repobrain with Docker?
Yes. Its repository has a Docker setup: a Dockerfile, a Compose file or a published image.
Also in MCP servers for developers
See all 86| Rank | Project | Score |
|---|---|---|
| 1 | codegraphLocal code knowledge graph that coding agents query over MCP | 96 out of 100 |
| 2 | gitnexusCode knowledge graph exposed to coding agents through MCP tools | 96 out of 100 |
| 3 | serenaMCP server giving coding agents symbol-level code search, editing and refactoring | 91 out of 100 |
| #4 | n8n-mcpMCP server exposing n8n node docs, templates and workflow management to AI assistants | 90 out of 100 |
| #5 | context7MCP server and CLI that feed current library docs to coding agents | 89 out of 100 |