Loading ToolBench
Local stdio bridge to the hosted Musevate MCP server: generate video from a prompt, an image, or reference images across many AI video models.
Four tools with basic descriptions but critical gaps in schema documentation and parameter clarity. Tool descriptions are present (10-50 chars) but lack actionable context for LLM selection. Input schemas visible for quote_video and generate_video show good parameter structure, but check_balance and check_video have empty or minimal schemas. No output schemas documented. No error handling guidance. No tool annotations (readOnlyHint/destructiveHint). Naming is verb-forward (check_, quote_, generate_) but descriptions don't explain WHEN to use each tool or dependencies between them (e.g., quote_video before generate_video).
Check the account balance and available credits
Check the status of a video generation by its ID
Generate a video from a prompt, an image, or reference images
Get a quote for the cost of generating a video before committing to it
No output schemas documented for any tool. LLMs cannot plan downstream calls or extract required fields (e.g., generation ID from generate_video result).
Descriptions lack actionable context. 'Check the account balance and available credits' (50 chars) does not explain when to call this vs quote_video, or what the response contains. No dependency hints.
quote_video and generate_video accept optional imageUrl and referenceImageUrls but descriptions do not explain the mutual exclusivity or how to choose between them.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 51 | 2026-07-28+ | v2 |
No error handling guidance. Tools marked READ_ONLY and WRITE but no recovery hints (e.g., 'If balance insufficient, call quote_video to see cost before retrying').