Codebase intelligence for AI and humans: code health scores, auto-generated docs, git analytics, dead code detection, and architectural decisions via MCP.
Python aicode-complexitycode-healthcode-intelligencecode-qualitydead-codedeveloper-toolsdocumentationgit-analyticsmcpopen-sourcepythonrefactoringstatic-analysistechnical-debt
Search Example
Walk through repowise search — keyword lookup over the wiki, semantic
search when an embedder is configured, and symbol lookup by name. No LLM
key required for fulltext and symbol modes.
Prerequisites
- A git repository with an indexed wiki.
repowiseonPATH(uv tool install repowiseor from this repo:uv sync --all-packages).
cd /path/to/your-repo
repowise init --yes --no-prose
1. Full-text search (default)
repowise search "rate limiting"
repowise search "pipeline" --limit 5
Searches rendered wiki pages. Works immediately after index-only init.
2. Symbol search
repowise search "run_pipeline" --mode symbol
repowise search "AuthService" --mode symbol --limit 10
Resolves indexed symbols to file paths and line numbers.
3. Semantic search (optional)
Requires a configured embedder (set during init / reindex):
repowise search "how are errors handled" --mode semantic
Without embeddings, use --mode fulltext explicitly or rely on the default.
Workspace scoping
repowise search "auth" --repo backend # one repo in a workspace
repowise search "auth" --all # fan out across all repos
Smoke checklist
| Step | Expected |
|---|---|
repowise search "pipeline" --limit 3 |
Wiki hits with scores (or empty if no match) |
repowise search "<symbol>" --mode symbol |
Symbol table with file + line |
| No index present | Clear error directing you to run repowise init |
Related docs
- CLI:
repowise search - Quickstart
- MCP
get_answer(synthesized Q&A in editors)