Loading ToolBench
Remote MCP server for Anthropic's Connectors Directory. Converts videos into structured documentation via Docsie's API.
Server has 3 tools with clear naming (verb_noun pattern) and reasonable descriptions (100-200 chars). All tools have input schemas with type definitions. However, parameter descriptions are sparse or missing entirely. analyze_video lacks descriptions for 'quality' and 'language' parameters despite having defaults. check_job_status has no parameter description. Output schemas are not documented. Error handling guidance is absent. Naming is strong (list_*, analyze_*, check_*), but schema completeness and parameter documentation fall short of production baseline.
Submit a video URL for AI-powered documentation generation. Analyzes the video (screen recording, training video, demo, etc.) and generates structured documentation. Processing takes 2-15 minutes depending on video length. Requires a Docsie account with AI credits.
Check the status of a video-to-docs job and retrieve results. Call this after analyze_video to poll for progress and retrieve the completed documentation when ready.
List available documentation types for video-to-docs conversion. Returns a description of each doc type so you can help the user pick the right option before calling analyze_video.
analyze_video: 'quality' and 'language' parameters have defaults but lack descriptions explaining what each tier does or which languages are supported. LLMs cannot infer parameter semantics from names alone.
check_job_status: 'job_id' parameter has no description. LLMs need to know the format, source (returned by analyze_video), and what happens if the job is not found.
No output schemas documented for any tool. LLMs cannot plan downstream calls or extract required fields (e.g., job_id from analyze_video response) without knowing the response structure.
No error handling guidance. If analyze_video fails (invalid URL, insufficient credits, rate limit), the tool provides no recovery hints. LLMs cannot self-correct or suggest next steps.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | C | 63 | 2026-07-28+ | v2 |
analyze_video description mentions '2-15 minutes' processing time but does not explain polling behavior or timeout expectations. check_job_status should document retry intervals and terminal states (success, failed, expired).