Loading ToolBench
LangGraph Agent with MCP Adapters - integrates MCP servers with LangGraph agents for tool utilization
Server has 4 tools with basic descriptions and simple string parameters. However, critical gaps undermine quality: (1) Duplicate tool 'get_weather' appears twice with identical definitions, suggesting incomplete deduplication or copy-paste errors. (2) Descriptions lack LLM-optimization guidance, they explain what the tool does but not WHEN to use it or what prerequisites exist. (3) No input schemas are visible in the source code; parameter definitions are inferred from docstrings only. (4) No output schemas documented. (5) No error handling guidance, tools return mock data or assume success. (6) Parameters lack type constraints (enums, ranges, patterns). The 'timezone' parameter in get_current_time has a default but no validation or list of valid timezones. This is a typical community server with functional tools but insufficient structure for reliable agent use.
Get current time information for the specified timezone. This function returns the current system time for the requested timezone.
Get current weather information for the specified location. This function simulates a weather service by returning a fixed response. In a production environment, this would connect to a real weather API.
Get current weather information for the specified location. This function simulates a weather service by returning a fixed response. In a production environment, this would connect to a real weather API.
Retrieves information from the document database based on the query. This function creates a retriever, queries it with the provided input, and returns the concatenated content of all retrieved documents.
Duplicate tool definition: 'get_weather' appears in both mcp_server_local.py and mcp_server_remote.py with identical signatures. This creates ambiguity for agents and suggests incomplete server consolidation.
No input schemas visible in source code. Tool parameters are documented only in docstrings; FastMCP decorator does not show explicit JSON Schema registration.
No output schemas documented. Tools return strings (get_weather, get_current_time) or concatenated text (retrieve) with no structured field definitions. Agents cannot plan downstream calls or extract typed data.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 46 | 2026-07-28+ | v2 |
Descriptions lack WHEN-to-use guidance and prerequisites. E.g., 'retrieve' does not explain that it requires a PDF at 'data/sample.pdf' or that it creates a new retriever per call (performance issue noted in comment but not in description).
No error handling or recovery guidance. Tools assume success (mock weather, file exists, timezone valid). No error messages tell agents what to do if a call fails (e.g., 'Invalid timezone. Valid timezones: Asia/Seoul, America/New_York, ...').