Loading ToolBench
A TypeScript-based MCP server that routes LLM requests to multiple backends (Ollama, llama.cpp, cloud) with support for external MCP tool calls, session management, embeddings, and advanced features like circuit breakers, rate limiting, and idempotency.
Single tool 'search' with minimal definition. Description is only 11 characters ('Search docs'), far below the 34-character minimum for A+ tools. No output schema documented. Tool appears to be a fixture/test artifact rather than a production-grade definition. Cannot verify actual tool registration or implementation details from provided source.
Search docs
Tool description is 11 characters ('Search docs'), violating the 10 - 1024 character guideline and falling far below the 34-character p10 baseline. LLMs cannot determine when or why to select this tool.
Input parameter 'q' lacks a description. LLMs cannot infer that 'q' means 'query string' or what format/length is expected.
No output schema documented. LLMs need to know what fields to expect in the response to plan downstream calls and extract data correctly.
Tool name 'search' is generic and lacks a noun. 'search_docs') is preferred. 'search' alone is ambiguous, search what? Users? Docs? Code?
Inferred effective spec: 2026-07-28+.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 28 | 2026-07-28+ | v2 |
Tool definition appears to be a test fixture (router/tests/fixtures/mcp-server.ts), not a production implementation. No evidence of actual tool registration, error handling, or integration with the MCP server runtime.