Loading ToolBench
Two tools with adequate naming (verb_noun pattern) and complete input schemas. However, descriptions lack LLM-optimization guidance (no WHEN to use, no prerequisites, no recovery hints). Parameter descriptions are present but minimal (10-30 chars). No output schema documentation, responses are free-text strings, forcing LLMs to parse unstructured data. Error handling returns raw API errors without recovery guidance. Missing tool annotations (readOnlyHint, idempotentHint). No pagination support despite potentially large result sets.
Search with Bocha AI Search, recognizes the semantics of search terms and additionally returns structured modal cards with content from vertical domains.
Search with Bocha Web Search and get enhanced search details from billions of web documents, including page titles, urls, summaries, site names, site icons, publication dates, image links, and more.
Output schema not documented. Both tools return free-text strings instead of structured JSON. LLMs must parse unstructured text, wasting tokens and increasing hallucination risk.
Descriptions lack WHEN-to-use guidance and recovery hints. 'Search with Bocha Web Search...' tells WHAT but not WHEN to prefer this over bocha_ai_search, or what to do if API key is missing.
Parameter descriptions are minimal (10-30 chars). 'Search query (required)' lacks format/length constraints. 'The time range...' lists options but doesn't explain when to use each (noLimit vs oneDay).
No pagination support. Results capped at 50 items, but no offset/limit parameters or next_cursor. Large result sets will be truncated without agent awareness.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 49 | 2026-07-28+ | v2 |
Error messages lack recovery guidance. 'Bocha Web Search API HTTP error occurred: 401 - Unauthorized' tells the agent nothing actionable. Should suggest 'Check BOCHA_API_KEY environment variable' or 'Verify API key at https://open.bochaai.com'.