Productivity-boosting RAG engine for codebases with multi-provider AI support and semantic search.
| Tool name | Description | Destructive? |
|---|---|---|
| ingest_codebase | Ingest a codebase directory into the DeepRepo vector store. Scans all supported files, chunks them, generates embeddings | ✓ no |
| query_codebase | Query the ingested codebase using RAG (Retrieval Augmented Generation). Embeds your question, finds the most relevant co | ✓ no |
| search_similar | Search for similar code chunks without using the LLM. Useful for finding related code snippets based on semantic similar | ✓ no |
| get_stats | Get statistics about the current DeepRepo vector store. Returns information about how many chunks are stored, the number | ✓ no |
| clear_history | Clear the conversation history in DeepRepo. Useful when you want to start a fresh conversation without context from prev | ✓ no |