search_memories
Search through saved conversation memories using vector similarity to find relevant information from previous interactions.
Instructions
Search the vector store for memories that match the query.
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes |
Implementation Reference
- server.py:39-58 (handler)The @mcp.tool() decorator registers the search_memories handler function, which performs a vector store search using OpenAI client and returns matching memory texts.
@mcp.tool() def search_memories(query: str): """Search the vector store for memories that match the query.""" vector_store = get_or_create_vector_store() print(vector_store.id) results = client.vector_stores.search( vector_store_id=vector_store.id, query=query, ) print(results) # Handle SyncPage response - iterate through the data content_text = [] for item in results.data: if hasattr(item, 'content'): for content in item.content: if content.type == "text": content_text.append(content.text) return {"status": "success", "results": content_text}