Loading ToolBench
Fast code intelligence engine for AI coding agents — single static binary MCP server
Single tool 'index_repository' has a basic description but lacks critical detail. Input schema is present with two parameters (repo_path, mode) but both lack descriptions, violating the requirement that every parameter must explain what it controls. The mode parameter accepts 'full' or 'incremental' but this is not declared as an enum in the schema. No output schema is documented. The tool description (47 chars) is below the 50-200 char LLM-optimized range and does not explain WHEN to use it, WHAT it returns, or any prerequisites. No error handling guidance. The tool performs a WRITE operation (indexing) but the description does not state this, leaving agents uncertain about side effects and retry safety.
Index a codebase repository to build a code intelligence graph
Parameters lack descriptions. 'repo_path' and 'mode' have no descriptions in the schema, violating pattern:tool-description. LLMs cannot infer that 'mode' must be 'full' or 'incremental'.
Mode parameter should be declared as an enum in JSON Schema, not free-form string. Current schema allows any string value, inviting hallucinated invalid modes.
No output schema documented. LLMs cannot plan downstream calls or extract results without knowing what fields the tool returns.
Tool description (47 chars) is too brief and lacks context. Does not explain WHEN to call it, WHAT it returns, or that it modifies state (WRITE risk). Should be 50-200 chars with actionable guidance.
Inferred effective spec: <=2025-11-25.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 42 | <=2025-11-25 | v2 |
No error handling guidance. If indexing fails (invalid repo path, permission denied, out of disk), the tool provides no recovery hints. LLMs cannot self-correct.