Loading ToolBench
A skill/tool for generating viral social media posts on specified topics
This MCP server has a single tool 'generate_post' with severe quality gaps. The tool has a valid name and basic schema, but the description is extremely sparse (52 characters), lacks context on when/why to use it, and provides no information about output structure or error handling. The parameter description is minimal. The actual implementation in client.py shows only a stub returning hardcoded output, providing no indication of real functionality. No error handling, no output schema documentation, no idempotency guarantees, and no security considerations are present. This server falls well below production quality standards.
Generate a viral social media post on a specified topic
Tool description is critically sparse (52 chars). Missing WHAT (beyond topic generation), WHEN to use it vs alternatives, output format, and any prerequisites. LLMs cannot reliably select or configure this tool.
Output schema is completely undocumented. The tool returns {post_text: string} but this is inferred from example usage, not declared. LLMs cannot plan downstream processing or validate responses.
Input parameter 'topic' has a trivial description ('The topic or subject for the social media post') that duplicates the parameter name. No constraints on format, length, or allowed values are documented. No guidance on realistic examples.
No error handling whatsoever. The stub implementation returns hardcoded output regardless of input. Real implementation must document failure modes (empty topic, API failures, rate limits) and provide recovery guidance.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 41 | 2026-07-28+ | v2 |
Tool implementation is a stub (hardcoded return value 'AI Agents are revolutionizing productivity!'). Cannot assess actual behavior, idempotency, or real error handling. Tool is non-functional for real use.
No idempotency guarantee. A 'generate' operation suggests non-deterministic output, if called twice with the same topic, should it return the same post or a different one? This must be documented.