菜单
Locate where a feature or symbol lives — match nodes by name, or pull the related cluster from a seed via spreading activation. Use to scope before reading.
用 Codex 或 Claude 帮你安装 复制这段 Prompt，粘贴到 Codex、Claude 或其他助手里，让它检查 Skill 页面并帮你完成安装。
基于 SOC 职业分类
Start from a failure, not a filename — turn a stacktrace, failing test, or diff into the exact code that matters, then read it. Use when you have an error and need to know where to look.
Blast radius before a change — everything a node affects (forward) or that depends on it (backward), ranked and located, with likely-dead code flagged. Also the one-hop view of direct connections.
Set up or rebuild the structural graph for a repo — on first use, after a big refactor, or to wire repo-graph into a project permanently.
Get oriented in a codebase before reading files — repo shape, entry points, where the graph is blind, and the full structural map. Use at the start of any task in an unfamiliar or large repo.
Follow the code across boundaries — a feature end-to-end, or the shortest path between two nodes (e.g. a frontend action to its backend handler). Links across the stack in one call instead of grepping both sides.
| name | repo-graph-find |
| description | Locate where a feature or symbol lives — match nodes by name, or pull the related cluster from a seed via spreading activation. Use to scope before reading. |
Find the relevant set, not one lucky file.
find with query=<text> → matching nodes with kind, qname, and path:line.find with query=<text> and expand=true → the related cluster ranked by relevance (Personalized PageRank from the matches).read, or run trace to see it end to end.Use the results to scope what you read. Don't read the whole repo to find one thing. (Debugging from an error rather than a keyword? Use the repo-graph-debug skill — find on the stacktrace, then read.)