repowise
Codebase intelligence for AI and humans: code health scores, auto-generated docs, git analytics, dead code detection, and architectural decisions via MCP.
- 7.9k stars
- 928 forks
- 391 open issues
- Language: Python
- License: AGPL-3.0
- updated
Install
local · uvx (python)claude mcp add repowise -- uvx repowise{
"mcpServers": {
"repowise": {
"command": "uvx",
"args": ["repowise"]
}
}
}{
"mcpServers": {
"repowise": {
"command": "uvx",
"args": ["repowise"]
}
}
}- Run the command above in your terminal (Claude Code), or paste the JSON config into claude_desktop_config.json (Claude Desktop).
- Replace any <placeholder> values with your API keys or paths.
- Restart Claude. The MCP server and its tools appear automatically.
Package name inferred from the repository name. Verify it exists on PyPI, or clone https://github.com/repowise-dev/repowise and follow its README.
Editor's note
Repowise is a Python package (installed via `pip install repowise`) that builds a local codebase intelligence index across five layers and exposes it to Claude Code and other MCP-compatible agents through nine task-shaped MCP tools. The five layers are: a tree-sitter dependency graph covering 15 languages with PageRank and Leiden community detection; git analytics including hotspots derived from churn-times-complexity, co-change pairs, bus factor, and reviewer suggestions; LLM-generated per-module documentation with hybrid RAG search combining full-text and vector retrieval; architectural decisions mined from eight sources and linked to graph nodes via relationship edges; and a code health score that calculates 1-to-10 file-level ratings from 25 deterministic biomarkers including McCabe complexity, LCOM4 cohesion, Rabin-Karp clone detection, and change entropy. The health scoring requires zero LLM calls and completes in under 30 seconds on a 3,000-file repository, with biomarker weights calibrated against a real defect corpus rather than hand-tuned. Engineering teams using Claude Code benefit most, gaining structured codebase context that reduces tool calls and file reads per query compared to agents working from raw source alone.
6 items in this repository
Use when a task asks why code is built a certain way, proposes architectural changes, compares implementation approaches, or mentions decision markers such as WHY, DECISION, TRADEOFF, or ADR in a Repowise-indexed repository.
Use when exploring, understanding, or answering questions about a Repowise-indexed codebase, including architecture, where code is implemented, how a module works, or which files are relevant before reading source.
Use when the user asks about unused code, cleanup, deleting files or exports, refactoring old areas, reducing bundle size, code hygiene, technical debt, or maintenance in a Repowise-indexed repository.
Use before modifying, refactoring, moving, or deleting files in a Repowise-indexed repository, especially shared utilities, core modules, public APIs, or files the user did not explicitly identify.
MCP Servers overview
What people ask about repowise
What is repowise-dev/repowise?
repowise-dev/repowise is mcp servers for the Claude AI ecosystem. Codebase intelligence for AI and humans: code health scores, auto-generated docs, git analytics, dead code detection, and architectural decisions via MCP. It has 7.9k GitHub stars and its last recorded update is dated 2026-10-23.
How do I install repowise?
You can install repowise by cloning the repository (https://github.com/repowise-dev/repowise) or following the README instructions on GitHub. ClaudeWave also provides quick install blocks on this page.
Is repowise-dev/repowise safe to use?
Our security agent has analyzed repowise-dev/repowise and assigned a Trust Score of 97/100 (tier: Verified). See the full breakdown of passed checks and flags on this page.
Who maintains repowise-dev/repowise?
repowise-dev/repowise is maintained by repowise-dev. The last recorded GitHub activity is dated 2026-10-23, with 391 open issues.
Are there alternatives to repowise?
Yes. On ClaudeWave you can browse similar mcp servers at /categories/mcp, sorted by popularity or recent activity.