Skip to content

armgit / letta

Public Imported from GitHub
0

Platform for stateful agents: AI with advanced memory that can learn and self-improve over time.

feat: add mcp tool simulator

jnjpng8b8536c3efd0
diff --git a/letta/server/rest_api/routers/v1/tools.py b/letta/server/rest_api/routers/v1/tools.py
index 864bbbf73..e841a64f6 100644
--- a/letta/server/rest_api/routers/v1/tools.py
+++ b/letta/server/rest_api/routers/v1/tools.py
@@ -794,6 +794,73 @@ async def generate_json_schema(
raise HTTPException(status_code=400, detail=f"Failed to generate schema: {str(e)}")
+# TODO: @jnjpng move this and other models above to appropriate file for schemas
+class MCPToolExecuteRequest(BaseModel):
+ args: Dict[str, Any] = Field(default_factory=dict, description="Arguments to pass to the MCP tool")
+
+
[email protected]("/mcp/servers/{mcp_server_name}/tools/{tool_name}/execute", operation_id="execute_mcp_tool")
+async def execute_mcp_tool(
+ mcp_server_name: str,
+ tool_name: str,
+ request: MCPToolExecuteRequest = Body(...),
+ server: SyncServer = Depends(get_letta_server),
+ actor_id: Optional[str] = Header(None, alias="user_id"),
+):
+ """
+ Execute a specific MCP tool from a configured server.
+ Returns the tool execution result.
+ """
+ client = None
+ try:
+ actor = await server.user_manager.get_actor_or_default_async(actor_id=actor_id)
+
+ # Get the MCP server by name
+ mcp_server = await server.mcp_manager.get_mcp_server(mcp_server_name, actor)
+ if not mcp_server:
+ raise HTTPException(
+ status_code=404,
+ detail={
+ "code": "MCPServerNotFound",
+ "message": f"MCP server '{mcp_server_name}' not found",
+ "server_name": mcp_server_name,
+ },
+ )
+
+ # Create client and connect
+ server_config = mcp_server.to_config()
+ server_config.resolve_environment_variables()
+ client = await server.mcp_manager.get_mcp_client(server_config, actor)
+ await client.connect_to_server()
+
+ # Execute the tool
+ result, success = await client.execute_tool(tool_name, request.args)
+
+ return {
+ "result": result,
+ "success": success,
+ }
+ except HTTPException:
+ raise
+ except Exception as e:
+ logger.warning(f"Error executing MCP tool: {str(e)}")
+ raise HTTPException(
+ status_code=500,
+ detail={
+ "code": "MCPToolExecutionError",
+ "message": f"Failed to execute MCP tool: {str(e)}",
+ "server_name": mcp_server_name,
+ "tool_name": tool_name,
+ },
+ )
+ finally:
+ if client:
+ try:
+ await client.cleanup()
+ except Exception as cleanup_error:
+ logger.warning(f"Error during MCP client cleanup: {cleanup_error}")
+
+
# TODO: @jnjpng need to route this through cloud API for production
@router.get("/mcp/oauth/callback/{session_id}", operation_id="mcp_oauth_callback")
async def mcp_oauth_callback(
diff --git a/letta/services/mcp_manager.py b/letta/services/mcp_manager.py
index 2c81a824c..4fc3ae266 100644
--- a/letta/services/mcp_manager.py
+++ b/letta/services/mcp_manager.py
@@ -287,7 +287,7 @@ class MCPManager:
@enforce_types
async def get_mcp_server(self, mcp_server_name: str, actor: PydanticUser) -> PydanticTool:
- """Get a tool by name."""
+ """Get a MCP server by name."""
async with db_registry.async_session() as session:
mcp_server_id = await self.get_mcp_server_id_by_name(mcp_server_name, actor)
mcp_server = await MCPServerModel.read_async(db_session=session, identifier=mcp_server_id, actor=actor)