build_vector_index
Create a vector index for semantic code search by generating embeddings with tree-sitter and Jina AI, enabling efficient and accurate querying of source code files.
Instructions
Build a vector index for semantic code search using tree-sitter and Jina embeddings
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| api_key | No | Jina AI API key for embeddings (can also be set via JINA_API_KEY env var) | |
| file_patterns | No | Optional list of glob patterns to filter files (e.g., ['**/*.py', '**/*.js']) | |
| force_rebuild | No | Force rebuild even if index already exists | |
| model | No | Jina embedding model to use | jina-embeddings-v3 |