Files
letta-server/letta/services/tool_executor/mcp_tool_executor.py
github-actions[bot] 940d648d42 fix(mcp): handle MCP tool errors gracefully to prevent Datadog alerts (#8687)
MCP tool errors (ToolError, McpError) are expected user-facing errors
from external MCP servers (e.g., "No connected account found"). These
were propagating through @trace_method decorator and being recorded
as errors in Datadog APM.

Changes:
- Add try/except to catch expected MCP errors in ExternalMCPToolExecutor
- Return ToolExecutionResult with status="error" instead of re-raising
- Log expected errors at INFO level instead of letting them trace as ERROR
- Remove stray 'pass' statement that was a no-op

Fixes #8685

🤖 Generated with [Letta Code](https://letta.com)

Co-authored-by: letta-code <248085862+letta-code@users.noreply.github.com>
Co-authored-by: datadog-official[bot] <datadog-official[bot]@users.noreply.github.com>
Co-authored-by: Kian Jones <11655409+kianjones9@users.noreply.github.com>
2026-01-19 15:54:43 -08:00

81 lines
3.2 KiB
Python

from typing import Any, Dict, Optional
from letta.constants import MCP_TOOL_TAG_NAME_PREFIX
from letta.log import get_logger
from letta.otel.tracing import trace_method
from letta.schemas.agent import AgentState
from letta.schemas.sandbox_config import SandboxConfig
from letta.schemas.tool import Tool
from letta.schemas.tool_execution_result import ToolExecutionResult
from letta.schemas.user import User
from letta.services.mcp_manager import MCPManager
from letta.services.tool_executor.tool_executor_base import ToolExecutor
from letta.utils import get_friendly_error_msg
logger = get_logger(__name__)
# MCP error class names that represent expected user-facing errors
# These are checked by class name to avoid import dependencies on fastmcp/mcp packages
MCP_EXPECTED_ERROR_CLASSES = {"McpError", "ToolError"}
class ExternalMCPToolExecutor(ToolExecutor):
"""Executor for external MCP tools."""
@trace_method
async def execute(
self,
function_name: str,
function_args: dict,
tool: Tool,
actor: User,
agent_state: Optional[AgentState] = None,
sandbox_config: Optional[SandboxConfig] = None,
sandbox_env_vars: Optional[Dict[str, Any]] = None,
) -> ToolExecutionResult:
mcp_server_tag = [tag for tag in tool.tags if tag.startswith(f"{MCP_TOOL_TAG_NAME_PREFIX}:")]
if not mcp_server_tag:
raise ValueError(f"Tool {tool.name} does not have a valid MCP server tag")
mcp_server_name = mcp_server_tag[0].split(":")[1]
mcp_manager = MCPManager()
# TODO: may need to have better client connection management
environment_variables = {}
agent_id = None
if agent_state:
environment_variables = agent_state.get_agent_env_vars_as_dict()
agent_id = agent_state.id
try:
function_response, success = await mcp_manager.execute_mcp_server_tool(
mcp_server_name=mcp_server_name,
tool_name=function_name,
tool_args=function_args,
environment_variables=environment_variables,
actor=actor,
agent_id=agent_id,
)
return ToolExecutionResult(
status="success" if success else "error",
func_return=function_response,
)
except Exception as e:
# Check if this is an expected MCP error (ToolError, McpError)
# These are user-facing errors from the external MCP server (e.g., "No connected account found")
# We handle them gracefully instead of letting them propagate as exceptions
if e.__class__.__name__ in MCP_EXPECTED_ERROR_CLASSES:
logger.info(f"MCP tool '{function_name}' returned expected error: {str(e)}")
error_message = get_friendly_error_msg(
function_name=function_name,
exception_name=e.__class__.__name__,
exception_message=str(e),
)
return ToolExecutionResult(
status="error",
func_return=error_message,
)
# Re-raise unexpected errors
raise