repowiserepowise

MCP server

repowise-dev / repowise

Codebase intelligence for AI and humans: code health scores, auto-generated docs, git analytics, dead code detection, and architectural decisions via MCP.

★ 8,166stars on GitHubPythonAGPL-3.0Curated

This is Repowise's own repository, indexed by Repowise. Every view in the sidebar is live, and it is what you get for your code. Index your repo free

github.com/repowise-dev/repowiseIndexed at ef827a2Up to date with upstreamRefreshes on every push

What is repowise?

repowise is a command-line tool and MCP server that indexes a code repository and its git history, so developers and coding agents can ask what a change touches. It builds the index once on your machine, and agents query it through the Model Context Protocol instead of re-reading files. Python is the primary language, and the repository also contains TypeScript packages. To install it, run git clone https://github.com/repowise-dev/repowise.git and follow the steps below. Its code health is 8.2 out of 10 (Good), ranked #45 of 4,000+ indexed repos. The last commit to repowise was on Nov 17, 2026, and 268 people have contributed.

As of commit 6ce6006, indexed Nov 17, 2026. What it is comes from a model reading the README; the rest is measured from the source and git history, not from the maintainer.

For agentsrepowise.md

Repo facts: type, license, health, rank, contributors
repowise facts
TypeMCP server
LicenseAGPL-3.0
Code health8.2 / 10 (Good)
Health rank#45 of 4,000+ indexed repos
Last commitNov 17, 2026
Contributors268
Bus factor1
Size7,348 files, 1,261,017 lines

Install repowise

Commands copied from the README

  1. 1Clone the repowise repository.
    git clone https://github.com/repowise-dev/repowise.git
  2. 2Move into the repowise folder.
    cd repowise
  3. 3Install repowise with uv.
    uv tool install repowise
  4. 4Change into the repository you want to index.
    cd /path/to/your/repo

Releases on GitHub (latest v0.61.0, Nov 13, 2026)

Full install notes on GitHub

How to use repowise

  1. 1Index the repository without an API key.
    repowise init --yes --no-prose
  2. 2Start the dashboard and the MCP server.
    repowise serve

Ask about repowise

Answers come from repowise's README and source, with the files they cite. Two free answers a day without an account.

Featured answer

repowise is a command-line tool and MCP server that indexes a code repository and its git history, so developers and coding agents can ask what a change touches. The index is built once on your machine, and agents query it through the Model Context Protocol instead of re-reading files. Input enters through packages/cli/src/repowise/cli/main.py, which runs indexing, documentation generation, and health and risk reports. The indexing engine in packages/core is called by the CLI commands, the server routes, and the MCP server in packages/server/src/repowise/server/mcp_server, which exposes the same index to agents as named tools. The file health scores come from packages/core/src/repowise/core/analysis/health/engine.py, so start reading with main.py, then the core package, then the MCP server.

Files per layer in repowise10 layers.

Basis: Inferred · knowledge graph· as of 6ce6006

Want this for your own codebase? Map your repo free

About this page

Now see it on your own codebase

Everything on this page, and every view in the sidebar, is built from a repository's source and git history. Point Repowise at yours and get the same map: module docs, the dependency graph, per-file health, hotspots and owners, and an index your coding agent can query over MCP.

Free for 2 public repos, with a page like this one. No card. See how Repowise works

Prefer local? pip install repowise && repowise init indexes your own checkout, no account needed. Star Repowise on GitHub

repowise-dev/repowise FAQ

What is repowise?

repowise is a command-line tool and MCP server that indexes a code repository and its git history. Developers and coding agents use the index to ask what a change touches, check code health and find dead code.

How does repowise work?

repowise builds an index of your code, git history, tests and decisions once, on your machine. Agents query that index through the Model Context Protocol, and developers read the same answers in a browser dashboard.

How do I install repowise?

repowise is installed with uv tool install repowise, and the README also lists pipx and pip as options. After changing into your repository, repowise init --yes --no-prose builds the index without an API key. See the install steps

What language is repowise written in?

repowise is written mainly in Python. The repository also contains TypeScript packages such as api-client and ui.

How is the repowise code organized?

The engine that indexes code lives in packages/core, and the repowise command line lives in packages/cli. packages/server holds the MCP server and HTTP routes, and the tests directory holds the test suite.

When should I use repowise?

Use repowise when you want to know what a change touches before you merge it, or when a coding agent keeps re-reading your codebase. It reports change risk, the tests to run, health scores and dead code.

What license does repowise use?

repowise is licensed under AGPL-3.0. The license identifier is AGPL-3.0 as listed for the repository.

Use repowise-dev/repowise in your agent

Point Claude Code, Cursor, Codex or VS Code at this repository. Your agent gets the index this page is built from: cited answers, callers, history and health for any file.

Run it in a terminal, then type /mcp in Claude Code and sign in with Repowise. No key to paste.

Docs
bash
claude mcp add --transport http repowise https://api.repowise.dev/mcp/repowise-dev/repowise
Claude Code, Claude, Cursor and VS Code sign you in on first use, free. Codex and Windsurf need an API key: sign in to create one.Sign in for a key

This repo is public, so anyone with a free Repowise account can connect.

Try asking

  • get_risk(["packages/core/src/repowise/core/analysis/health/dataflow/slice.py"])
  • get_answer("how is repowise put together?")
  • get_why("why is repowise structured this way?")

Key features of repowise

  • Builds a call graph across 28 languages and generates a wiki page for every module and file.
  • Scores every file from 1 to 10 using 53 deterministic detectors covering defect risk, maintainability and performance.
  • Scores any commit or branch from 0 to 10 for change risk against the repository's recent commits, and lists the tests to run.
  • Reports dead code grouped by confidence tier through the repowise dead-code command.
  • Mines decisions from ADRs, comments, commits, pull requests and agent sessions, and checks docs for drift against the code.
  • Gives Claude Code, Codex, Cursor and any MCP host the same index through MCP tools.

How repowise-dev/repowise works

Input enters through the repowise command line in packages/cli/src/repowise/cli/main.py, which runs indexing, documentation generation, and health and risk reports. The indexing engine lives in packages/core, and the CLI commands, the server routes and the MCP server all call into it. The health analysis engine in packages/core/src/repowise/core/analysis/health/engine.py produces the file health scores. The MCP server in packages/server/src/repowise/server/mcp_server exposes the same index to agents as named tools. The hook file packages/cli/src/repowise/cli/augment_hook.py is one of the entry points that AI coding agents call during their work. The scripts directory holds benchmark and check scripts that validate the knowledge graph of files and import edges.

The codebase has 7,349 files and 95,052 symbols in 8 modules, led by Python, TypeScript and JavaScript. Code health is 8.2 out of 10, rated good.