framework-search

Semantic code discovery uses Qdrant vector search to index and retrieve framework code by intent. It indexes Python modules, hooks, and docs, allowing you to locate implementation details using natural-language descriptions of behavior or concepts.

Yes, broad framework coverage searches across Python code, skills, hooks, scripts, docs, and configuration files. You can ask where a concept is implemented and surface relevant files without requiring exact filenames or symbol names.

Keep semantic search results current using lazy mtime reindexing, file watcher logic, and git-hook refresh workflows. These mechanisms automatically update the Qdrant vector index when framework files are modified.

Yes, semantic code search supports operational maintenance tasks like refactoring impact analysis and onboarding. It helps locate the right place in large codebases by turning manual grep sessions into fast semantic lookups.

Yes, Qdrant vector search is required as the underlying engine for semantic code indexing and retrieval. The skill uses deterministic chunking and Qdrant to perform semantic lookup across the framework repository.

Use semantic code discovery to find implementation details from natural-language descriptions such as fallback logic, indexing flows, or hook behavior. It retrieves relevant framework files based on conceptual meaning rather than exact symbols.