# Editorial Prover MCP MCP > Editorial Prover runs a structured self-audit on any text to eliminate predictable AI patterns. It forces your agent to prove the content's rhythm, identify specific weaknesses, and ensure structural variety—making machine-written copy genuinely human. ## Overview - **Category:** productivity - **Price:** Free - **Tags:** editorial-quality, ai-writing, content-audit, humanlike-text, multi-language, structured-reasoning, decision-pivots, burstiness ## Description Your AI client writes great drafts, but sometimes the output sounds identical to every other draft. The sentences are uniform, the transitions are predictable, and the structure feels lifted from a textbook. Editorial Prover fixes that pattern problem at the source. Instead of just checking for grammar errors, this MCP forces your agent through five distinct thinking steps: figuring out who you're talking to, justifying the opening hook, varying sentence lengths, pinpointing the weakest line, and mapping paragraph structures to ensure variety. You connect Editorial Prover via Vinkius and run any text through its audit. The result isn't just a report; it's a forced act of deep editorial thinking that makes your final copy sound like a person wrote it—a real person. ## Tools ### audit_copy Runs a comprehensive self-audit on your text by checking for weak hooks, monotonous rhythm, repetitive structure, generic language, and filler phrases. ## Prompt Examples **Prompt:** ``` I'm writing a landing page headline for a CLI tool that automates database migrations. Audit my copy: 'A comprehensive tool for streamlined database migration management.' ``` **Response:** ``` Verdict: WEAK_HOOK. Your headline describes the feature ('comprehensive tool for streamlined management') instead of selling the benefit. It hedges with generic adjectives. Rewrite to commit: 'Stop breaking production with manual migrations.' — that's a position, a pain point, and urgency in 7 words. ``` **Prompt:** ``` I wrote a 4-paragraph product description in Portuguese. Every paragraph starts with 'Esta ferramenta...' — is that a problem? ``` **Response:** ``` Verdict: REPETITIVE. Starting 4 paragraphs with the same structure is the #1 marker of AI-generated text (Northeastern syntactic templates research, 2025). Map your structures: P1: problem→consequence, P2: 'Esta ferramenta' statement→evidence, P3: question→answer, P4: testimonial→CTA. Each paragraph must have a different architecture. ``` **Prompt:** ``` After auditing all 3 sections of my README, what's the editorial report? ``` **Response:** ``` Editorial Report — 3 sections audited: Section 1 (intro): PUBLISH_READY. Section 2 (features): MONOTONE → revised → PUBLISH_READY. Section 3 (getting started): PUBLISH_READY. Recommendation: CONTENT_APPROVED. Revision cycles: 1. Average burstiness CV: 0.47 (human-like rhythm). ``` ## Capabilities ### Determine Target Reader Desire Forces the AI to specify exactly who is reading the content and what they need from it. ### Validate Hook Commitment Checks if your headline takes a clear, assertive position instead of just describing the topic. ### Analyze Sentence Rhythm Ensures sentence lengths vary dramatically, mixing short punches with longer, descriptive explanations. ### Identify Weakest Content Points Locates the weakest sentence in any given section and requires a decision on whether to rewrite or cut it. ### Map Structural Variety Validates that adjacent paragraphs don't share the same basic architectural pattern (e.g., question-answer, claim-evidence). ## Use Cases ### The Technical Manual Needs a Human Touch A technical writer needs to update a product guide for cloud engineers. Instead of having the agent write 10 paragraphs that all sound like Wikipedia entries, they run the content through Editorial Prover. The tool forces structural variation and ensures the tone shifts between explanatory and punchy, making it feel less academic and more helpful. ### The Landing Page Headline is Too Bland A marketer writes a headline that merely states what their CLI tool does: 'A comprehensive solution for migration management.' The agent runs the copy through the MCP, gets flagged for a weak hook, and is forced to commit to a pain point instead: 'Stop breaking production with manual migrations.' ### The Blog Post Feels Repetitive A content creator writes a four-part series post where every paragraph starts identically. The MCP detects this structural repetition and demands the agent map out unique architectures for each section, ensuring the reader never gets bored with the format. ### The Pitch Deck Text Is Too Fluffy A startup founder types up a pitch deck description filled with 'synergy' and 'leverage.' The agent runs it through the tool, which strips out every filler phrase and forces the copy to use concrete language that gets straight to the point. ## Benefits - Boosts natural rhythm: The audit ensures you mix short, sharp sentences with longer explanations, eliminating the flat 15-20 word sentence pattern typical of LLMs. You won't sound monotonous again. - Eliminates filler words: It flags and removes corporate fluff like 'It is important to note' or 'furthermore,' forcing your copy to be direct and meaningful. - Varied structure, varied flow: Instead of every paragraph following the same claim-evidence pattern, this MCP maps out different architectures (question→answer, story→lesson) for a richer read. - Stronger hooks: The tool forces you to justify your opening line, ensuring it makes an immediate, committal promise instead of just vaguely describing the content. - Audience focus: By demanding you name the specific reader and their desires, the process prevents vague writing aimed at 'everyone' which means no one. ## How It Works The bottom line is that it forces your AI agent to act like an experienced human editor, not just a content generator. 1. Feed your draft content into this MCP and trigger the `audit_copy` tool. 2. The system executes a five-point self-audit, forcing the connected AI client to reason through audience needs, hook strength, rhythm variation, structural mapping, and filler word purging. 3. You receive a structured report detailing which patterns failed (e.g., 'Monotone Rhythm detected') alongside specific recommendations for rewriting. ## Frequently Asked Questions **How does Editorial Prover MCP improve my writing quality?** It improves quality by making your AI agent reason about its own choices. It forces it to justify hooks, vary rhythm, and map structural differences, which is the act of thinking itself. **Can Editorial Prover MCP handle non-English content?** Yes, the audit process is language-agnostic. The core pivots—like identifying audience desire or mapping structure—are universal editorial questions that apply to any text. **What does `audit_copy` actually detect in my writing?** `audit_copy` looks for five specific weaknesses: weak hooks, monotonous rhythm, structural repetition, generic vocabulary, and filler phrases. It tells you exactly where the copy fails to sound human. **Do I need Editorial Prover MCP for technical manuals?** If your manual needs to be engaging—not just factual—you should use it. The tool helps prevent repetitive, dull sections by forcing structural variety across different chapters or topics. **How does Editorial Prover MCP integrate with my existing AI client?** The MCP connects using standard protocols, ensuring compatibility across any agent that supports Vinkius. You simply connect your preferred AI client—like Cursor or Claude—through the Vinkius catalog to gain instant access. **Is the content I send to `audit_copy` secure and private?** Yes, all text processed by this MCP remains confidential. We only use your input for the audit; your data is not stored or used to train any external models. **What happens if I send very long documents to `audit_copy`?** The tool manages large inputs by breaking them into manageable sections for analysis. If a document exceeds the maximum token limit, you'll receive an error specifying which section needs to be processed next. **Are there any rate limits when I use Editorial Prover MCP frequently?** Vinkius manages resource allocation across all users to ensure system stability. While continuous, high-volume usage may require a subscription upgrade, the platform is built for sustained throughput. **Does Editorial Prover detect AI-generated text?** No — and that's intentional. Detectors catch problems after the fact. Editorial Prover prevents them at the source by forcing the agent to think like an editor BEFORE writing. The only server-side check is burstiness (sentence length variance), which is a language-agnostic mathematical validation, not a detection algorithm. **Does it work for languages other than English?** Yes — every language. The 5 Decision Pivots are universal editorial questions (who is the reader? does the opening grab? which sentence is weakest?) that work regardless of language. The burstiness check splits on sentence-ending punctuation (periods, question marks, exclamation marks — including CJK equivalents) and measures word count variance, which is pure math. There are no English-specific blocklists or grammar rules. **Why does the agent send its own verdict instead of the tool computing it?** Because the commitment IS the thinking. If the server computed the verdict automatically, the agent would fill in fields mechanically without reasoning about the outcome. By forcing the agent to declare 'I believe this is PUBLISH_READY' and then validating that declaration against the pivots, the agent must actively reason about whether its editorial self-assessment is consistent. This is the same pattern used by Sequential Thinking — the LLM decides when it has thought enough, which is what makes it think more.