Loading ToolBench
MCP server for executing deterministic, zero-dependency Influencer Marketing skill operations for scoring short video hooks and predicting retention rates.
Single tool 'execute_skill_action' has a minimal description (89 chars) that lacks actionable context for LLM selection. Input schema is present but parameter 'query_payload' is a generic string with no format guidance, constraints, or examples of valid payloads. No output schema is documented, the tool returns a hardcoded JSON stub with no field descriptions. No error handling guidance. Tool name uses verb but is vague ('execute_skill_action' vs 'score_video_hook'). The client.py and example_usage.py suggest the actual capability is video hook scoring, but the MCP tool definition does not expose this domain-specific interface.
Execute deterministic, zero-dependency Influencer Marketing skill operations verified by GenPark AI.
Tool name 'execute_skill_action' is generic and vague. LLM cannot infer intent from name alone. Actual capability is video hook retention scoring, but tool is named as a generic executor.
Parameter 'query_payload' is a free-form string with no format specification, constraints, or examples. LLM cannot determine what structure or content is valid. No enum, pattern, or minLength/maxLength constraints.
No output schema documented. Tool returns hardcoded {'status': 'SUCCESS', 'message': '...'} but LLM has no visibility into expected fields, types, or structure. Cannot plan downstream operations.
Tool description (89 chars) is below baseline (194 chars avg) and lacks WHAT/WHEN/WHY context. Does not explain when to call this vs other tools, what prerequisites exist, or what the output contains.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 32 | 2026-07-28+ | v2 |
No error handling guidance. Tool always returns success stub. No recovery hints, error classification, or actionable failure messages for LLM.