v0.2.0

Added

• Smart Type Inference: When a type filter returns no matches, the search now automatically falls back to inferred types or performs a type-less search to surface relevant results.
• Cached Search Index: Build and cache a derived search index to speed up suggestions and improve search quality.
• New suggest_python_doc_types Tool: A dedicated tool to suggest documentation types based on query context.
• Type Suggestions: Enhanced python_docs tool now provides type suggestions when exact matches aren't found.
• Schema-Aware Cache Invalidation: Cached documents missing required fields (e.g., anchorIndex) are automatically refetched even within TTL.

Enhanced

• Search Quality: Tightened type matching to require exact type equality for more accurate results.
• Result Display: Updated formatSearchResults to display fallback information and type suggestions when applicable.
• DocService API: New methods getSearchIndex(), searchWithFallback(), suggestTypes(), and getAvailableTypes().

Technical

• Added SearchIndex and dedicated cache path to CacheManager.
• Added comprehensive tests for search index inference.
• Added tests for TTL-based fetch and schema-aware invalidation.
• Added tests for garbage collection of expired cache files.

Documentation

• Updated README to document schema-aware invalidation behavior.

Full Changelog: https://github.com/yriveiro/opencode-python-docs/commits/v0.2.0