semantic-searchreindex
Use when you want to rebuild or refresh the semantic search index for the codebase
When & Why to Use This Skill
This Claude skill enables the rebuilding and refreshing of a codebase's semantic search index, ensuring the AI agent maintains an accurate, high-fidelity understanding of the project. By synchronizing the search index with the latest code changes, it significantly improves the relevance of code retrieval, debugging accuracy, and the quality of AI-generated architectural insights.
Use Cases
- Case 1: Post-Refactoring Update - After performing significant structural changes or renaming multiple files, use this skill to ensure the AI doesn't reference deprecated code paths.
- Case 2: Fresh Project Onboarding - When first introducing a large, complex repository to Claude, trigger a full reindex to establish a comprehensive semantic map for better navigation.
- Case 3: Index Troubleshooting - If the AI provides inconsistent or 'hallucinated' code references, use the 'clean' option to wipe and rebuild a fresh index from scratch for improved reliability.
- Case 4: Continuous Development Sync - Periodically refresh the index after merging multiple Pull Requests to keep the AI's context perfectly aligned with the current master branch.
| name | semantic-search:reindex |
|---|---|
| description | Use when you want to rebuild or refresh the semantic search index for the codebase |
Semantic Search Reindex
Trigger a full reindex of the codebase for semantic search.
Action
- Call
mcp__semantic-search__reindextool withforce: true - Immediately call
mcp__semantic-search__get_statusto show initial progress - Inform the user that indexing is running in the background
- Tell them they can check progress with
/semantic-search:status
Options
If the user says "clean" or "fresh", use clear_first: true to wipe the index before rebuilding.