fast-repo-context

Semantic code search retrieves relevant code snippets from your repository by intent rather than keywords. Fast-repo-context performs this search on a local git repository using a comprehensive code snapshot, returning structured JSON output with code location and context to accelerate feature implementation and debugging.

Use semantic code search to systematically retrieve only the relevant repository context you need. This approach avoids flooding your context window while enabling rapid analysis of authentication modules, feature implementations, and architectural patterns across distributed files.

Yes, code search supports optional filters including --glob for file patterns, --context for scoping, and --limit for result count. These filters let you narrow searches to specific directories or file types, returning only code snippets that match your query intent.

Code search optionally augments results with knowledge graph queries to provide semantic and architectural context beyond code-level insights. This complements structural analysis with higher-level understanding of how components relate and interact.

You need a local git repository and a fresh repomix-output.xml snapshot (a comprehensive code inventory). The skill enforces freshness checks and structured workflow steps to ensure search results reflect current code state.

Semantic search finds code by behavior and intent, not literal strings, making it faster for locating implementations across unfamiliar codebases. It returns structured JSON with location and context, avoiding false matches and overwhelming noise from keyword-only approaches.