PinRAG

Instructions

Returns unique document IDs (PDF file names, video IDs, discord-alicia-1200-pcb, owner/repo/path for GitHub, etc.)
currently in the vector store, plus total chunk count. Uses server config
for vector store location and collection.

Args:
    tag: Optional tag to filter: only list documents that have this tag.
    ctx: MCP request context (injected by the server; unused).

Returns:
    Dictionary containing documents, total_chunks, persist_directory,
    collection_name, document_details.

Implementation Reference

• src/pinrag/mcp/server.py:242-274 (handler)

  The implementation of list_documents_tool which calls list_documents function in a separate thread.

  async def list_documents_tool(
      tag: Annotated[
          str,
          Field(
              description="Optional tag to filter: only list documents that have this tag."
          ),
      ] = "",
      ctx: Context | None = None,
  ) -> dict:
      """List all indexed documents in the PinRAG index.
  
      Returns unique document IDs (PDF file names, video IDs, discord-alicia-1200-pcb, owner/repo/path for GitHub, etc.)
      currently in the vector store, plus total chunk count. Uses server config
      for vector store location and collection.
  
      Args:
          tag: Optional tag to filter: only list documents that have this tag.
          ctx: MCP request context (injected by the server; unused).
  
      Returns:
          Dictionary containing documents, total_chunks, persist_directory,
          collection_name, document_details.
  
      """
  
      def _run() -> dict:
          return list_documents(
              persist_dir=config.get_persist_dir(),
              collection=config.get_collection_name(),
              tag=tag or None,
          )
  
      return await anyio.to_thread.run_sync(_run)