MCP Server Memos

Instructions

Implementation Reference

• src/mcp_server_memos/server.py:95-107 (handler)

  The handler function in MemoServiceToolAdapter that validates input, calls the memos API to list memos matching the keyword, and returns the search results as text content.

  async def search_memo(self, args: dict) -> list[types.TextContent]:
      try:
          params = SearchMemoRequest.model_validate(args)
      except Exception as e:
          raise McpError(types.INVALID_PARAMS, str(e))
  
      req = memos_api_v1.ListMemosRequest(
          filter=f"row_status == 'NORMAL' && content_search == ['{params.key_word}']"
      )
      res = await self.memo_service.list_memos(list_memos_request=req)
      content = ", ".join([memo.content for memo in res.memos])
      content = f"Search result:\n{content}"
      return [types.TextContent(type="text", text=content)]

• src/mcp_server_memos/server.py:29-39 (schema)

  Pydantic BaseModel defining the input schema for the search_memo tool, with a required 'key_word' field.

  class SearchMemoRequest(BaseModel):
      """Request to search memo"""
  
      key_word: Annotated[
          str,
          Field(
              description="""The key words to search for in the memo content.""",
          ),
      ]

• src/mcp_server_memos/server.py:160-164 (registration)

  Registration of the 'search_memo' tool in the list_tools() decorator, specifying name, description, and input schema.

  types.Tool(
      name=MemosTools.SEARCH_MEMO,
      description="Search for memos",
      inputSchema=SearchMemoRequest.model_json_schema(),
  ),

• src/mcp_server_memos/server.py:185-186 (registration)

  Dispatch logic in the call_tool() decorator that routes calls to the search_memo handler when the tool name matches.

  if name == MemosTools.SEARCH_MEMO:
      return await tool_adapter.search_memo(args)

• src/mcp_server_memos/server.py:15-15 (helper)

  Constant definition for the tool name within the MemosTools enum.

  SEARCH_MEMO = "search_memo"