Loading ToolBench
MCP server that creates AI-enhanced posts on Twitter/X with content enhancement powered by Google Gemini API
This server has 3 tools with inconsistent quality. One tool (addTwoNumbers) is a placeholder that has no business purpose. The Twitter tools (createPost, previewEnhancedContent) have adequate descriptions but critically lack parameter descriptions for key inputs. No output schemas are documented. Error handling is minimal and does not guide LLM recovery. The server mixes a trivial math tool with domain-specific Twitter operations, violating single-responsibility principles.
Add two numbers
Create an AI-enhanced post on X (formerly Twitter). The content will be automatically enhanced with engaging language, emojis, and relevant hashtags.
Preview how your content will be enhanced by AI without posting to Twitter
Missing or null parameter descriptions in critical tools. The addTwoNumbers tool shows 'description': null for inputs. Critical parameters like 'status' in Twitter tools lack formal description documentation visible in the actual tool registration, violating the requirement that 100% of A+ tool parameters have descriptions.
No documented output schemas. Callers (and the LLM) do not know what fields to expect from tool responses. All three tools return content arrays with text, but this structure is not formally specified, forcing the LLM to infer response format.
Input parameters lack constraints. The 'status' parameter should specify Twitter's 280-character limit, forbidden characters, and format. Without constraints, the LLM may generate oversized or malformed content. No min/max, enum, or pattern validation is documented.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | F | 47 | 2026-07-28+ | v2 |
| 2026-03-09 | F | 47 | - | v1 |
Error handling does not guide recovery. When the Twitter API fails, the tool returns 'Error: <message>' but does not tell the LLM whether to retry, ask the user, or abandon the operation. No error categorization (retryable, user-fixable, fatal).
Tool composition violation: addTwoNumbers is a generic math utility in a Twitter server. Tools should each do one thing and serve a cohesive domain. This mixing signals poor API design and wastes context on irrelevant operations.
No idempotency or confirmation pattern. createPost is a destructive write operation that posts to Twitter. There is no dry-run, preview confirmation, or way to prevent accidental posts. An LLM could trigger multiple posts on retry.
Credentials (Twitter API keys) are passed via environment variables but are visible in the code. The .env loading pattern is standard, but there is no mention of secret rotation, least-privilege scoping, or audit logging of API calls.