AI text humanizer MCP — remove AI patterns, rewrite for platforms, detect AI tells
Four tools with complete input schemas and descriptions, but significant gaps in output documentation and error handling. Tool names follow verb_noun convention (humanize_text, detect_ai_patterns, rewrite_for_platform). Descriptions are adequate (100-200 chars) but lack context on when to use each tool vs. alternatives. Parameters are well-typed with enums for constrained inputs (tone, context, platform). Critical weakness: no documented output schemas, no error recovery guidance, no pagination for batch results, and no indication of which operations are idempotent or destructive. Batch tool lacks per-item error reporting. Platform-specific tool omits guidance on author_voice parameter usage.
Scan text for AI-generated patterns and phrases. Returns a list of specific issues found and a human-ness score.
Humanize multiple short texts at once — great for social media posts, email subject lines, or ad variations.
Rewrite AI-generated text to sound natural and human. Removes robotic patterns, filler phrases, and AI tells.
Rewrite content specifically optimized for a platform: Twitter/X, LinkedIn, Reddit, cold email, or Product Hunt.
No documented output schemas. Tools return unstructured text via LLM; LLMs cannot plan downstream calls or extract structured data. Violates pattern:response-shaper.
No error handling or recovery guidance. If Groq API fails, LLM receives raw error with no actionable next step. Missing pattern:recovery-guide.
humanize_batch lacks per-item error reporting. If 1 of 10 texts fails, entire batch fails with no indication which item caused the problem.
No tool annotations (readOnlyHint, idempotentHint). All tools are read-only but this is not declared. Agents cannot reason about retry safety or side effects.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | D | 52 | 2026-07-28+ | v2 |
author_voice parameter in rewrite_for_platform is optional but lacks guidance on format or impact. Description does not explain when/why to use it.