Docs
Your repoas a graph.
Install, map the repository, query it from the terminal or over MCP. Five minutes, all of it local.
Install
Two commands.
One installs it, the other registers it with your assistants.
$ uv tool install graphifyy
installed graphify
# or: pipx install graphifyy · pip install graphifyy
$ graphify install
detected Claude Code, Cursor, GitHub Copilot
/graphify skill registered · 17 assistants supportedBuild
Point it at the repository.
One command in your assistant, and a flag to re-scan what changed.
# in your assistant (Claude Code, Cursor, Copilot and the rest):
/graphify .
graphify-out/graph.html interactive graph
graphify-out/GRAPH_REPORT.md architecture report
graphify-out/graph.json machine-readable graph
# re-scan only what changed: /graphify . --update
# a deeper multi-pass read: /graphify . --mode deepAsk
Three verbs.
Query what connects, path between two, explain why an edge exists.
$ graphify query "what connects auth to the database?"
AuthService -> SessionStore -> DatabasePool [EXTRACTED]
src/auth/service.py:42 -> src/db/pool.py:17
$ graphify path "UserService" "DatabasePool"
UserService -> UserRepository -> DatabasePool (2 hops)
$ graphify explain "RateLimiter"
RateLimiter · class · src/middleware/rate_limit.py
called by ApiGateway, WebhookHandler · calls RedisClientServe
Hand it to an assistant.
The graph over MCP, and the pull request queue from the terminal.
$ python -m graphify.serve graphify-out/graph.json
MCP server ready (stdio) · 10 tools
$ graphify prs --triage
#482 HIGH overlaps #479 on RedisClient · review first
#479 LOW isolated changeThe server reads the graph and nothing else. It writes no code and touches no files, so the worst an assistant can do is ask a question you did not expect.
Next
Where to go from here.
The command reference, the server's tools, and the thinking under both.