humanizer-lint

Humanizer Lint

Audit, lint, and humanize AI-generated prose across 100+ languages using deterministic local stylometric math before returning text to the user.

• Eliminates robotic LLM writing patterns when users ask to "humanize text", "remove AI slop", "fix uniform sentence rhythm", or "write natural blog, copywriting, SEO, and documentation prose"
• Provides mandatory node <skill_dir>/scripts/analyze.mjs --stdin --json verification and npx humanizer-lint CLI to enforce Burstiness (CV >= 0.45), Adjacent Rhythm Delta, and Adaptive MATTR vocabulary diversity
• Ships with 8 deep anti-slop lexical & morphological rule packs (en, id, es, ms, fr, de, pt, zh) plus Universal Stylometric Mode (Intl.Segmenter) for 100+ languages with zero external dependencies
• Supports personal author voice cloning via node <skill_dir>/scripts/fingerprint.mjs and passes all 3 skills.sh security audits (Gen Agent Trust Hub, Socket, Snyk) with zero network calls

When to use

• Humanize agent-generated drafts, articles, essays, copywriting, and documentation in any language (--lang auto).
• Audit and fix uniform sentence rhythm (low Burstiness $CV$ and monotonous 3-sentence windows).
• Strip LLM stock phrases, filler transitions, and rigid translationese morphology (EN, ID, ES, MS, FR, DE, PT, ZH).
• Match a user's personal writing style using a local voice fingerprint profile (scripts/fingerprint.mjs).
• Verify edited prose deterministically (PASS / FAIL) before returning it to the user.

Supported Languages (Two-Tier Multilingual Architecture)

• Tier 1 — Full Lexical Anti-Slop + Stylometric Math (8 Languages): en (English), id (Bahasa Indonesia), es (Español), ms (Bahasa Melayu), fr (Français), de (Deutsch), pt (Português), zh (中文 / Mandarin).
• Tier 2 — Universal Stylometric Engine (100+ Languages via Intl.Segmenter): Automatic sentence-length Burstiness ($CV$), Adjacent Rhythm Delta ($\Delta_{adj}$), Adaptive $MATTR$, paragraph symmetry, starter repetition, and compression entropy for Latin, Cyrillic, Arabic, Japanese, Korean, Hindi, Turkish, Vietnamese, and other scripts via --lang auto or --lang universal.

Agent-SEO Trigger Intents & Search Keywords

• Core Humanizer & Anti-Slop: humanizer, ai humanizer, humanize text, anti-slop, remove ai slop, prose linter, stylometric linter, bypass robotic ai writing, burstiness checker, perplexity and burstiness, mattr lexical diversity.
• Writing, Copywriting & Docs Workflows: writing assistant, copywriting linter, blog post writer, seo article humanizer, markdown prose linter, technical documentation editor, proofreading and editing, natural translation polishing, voice cloning stylometry.
• Indonesian & Malay Intents: humanizer bahasa indonesia, cara agar tulisan ai tidak terdeteksi, perbaiki tulisan ai yang kaku, ubah teks chatgpt jadi natural, cek skor tulisan ai, parafrase anti-ai, semak ayat ai bahasa melayu.
• Global Multilingual Intents: humanizador de texto ia (ES/PT), humaniseur de texte ia (FR), ki-text humanisieren (DE), 消除AI味 / AI文章人性化 (ZH).

Do not use it to alter code, commands, structured data, quotations, legal language, or factual identifiers. The sanitizer excludes common Markdown structures, but the agent must still preserve meaning and supported facts.

Prerequisites

• Node.js 18.3 or newer.
• Resolve this skill's installed directory as <skill_dir> at runtime.
• Optional custom profile: <skill_dir>/profiles/custom-voice.json.

Closed-loop protocol

1. Resolve context

1. Resolve <skill_dir> from this SKILL.md; never assume the current working directory.
2. Determine auto, universal, or a two-letter ISO 639-1 code. auto selects a bundled pack when evidence is strong; otherwise it uses universal.
3. If <skill_dir>/profiles/custom-voice.json exists, pass it with --profile. Otherwise use the bundled default profile.
4. Keep source facts, names, numbers, dates, quotations, citations, and technical terms unchanged unless the user explicitly requests a factual edit.

Completion criterion: language, analyzer path, and optional profile path are known before drafting.

2. Draft with asymmetric rhythm

Write the first draft with concrete subjects and verbs. Mix short 2 to 6 word sentences with longer 26 to 38 word sentences where the meaning supports them. Avoid generic introductions and summary paragraphs that merely repeat earlier content.

Do not distort meaning to hit a metric. Sentence-length variation must come from legitimate splitting, merging, or reordering of supported material.

Completion criterion: the draft preserves every source claim and contains no invented detail.

3. Run deterministic verification

Prefer stdin so the target repository remains clean:

node <skill_dir>/scripts/analyze.mjs --stdin --lang auto --json

Pipe the complete draft to stdin. With a custom profile:

node <skill_dir>/scripts/analyze.mjs --stdin --lang auto --profile <skill_dir>/profiles/custom-voice.json --json

Never skip this command. Read status, score, metrics, and every item in violations.

Completion criterion: a parseable JSON report exists for the exact draft being considered.

4. Correct only flagged prose

If status is FAIL:

1. Group overlapping violations by paragraph. Whole-text statistical diagnostics (LOW_BURSTINESS, MISSING_SENTENCE_RANGE, LOW_ADJACENT_DELTA, and LOW_LEXICAL_DIVERSITY) identify the global gate when no single sentence is solely responsible.
2. Edit only the cited sentences or paragraphs, or make the smallest whole-text rhythm correction named by a statistical diagnostic.
3. Follow each fixInstruction without adding or dropping facts.
4. Run the analyzer again on the complete revised draft.
5. Stop after three analyzer iterations.

If status is SKIP, statistical stylometry is unavailable. Fix any lexical violations that are still reported, preserve the short form, and state that the statistical gate was skipped because the text is too short.

If the third iteration still returns FAIL, do not present the draft as verified. Return the remaining diagnostics and ask whether the user wants a broader rewrite.

Completion criterion: return prose only after PASS, or an explicitly labeled short-text SKIP with no unresolved lexical violations.

Analyzer commands

node <skill_dir>/scripts/analyze.mjs --file draft.md --lang id --json
node <skill_dir>/scripts/analyze.mjs --file draft.md --lang en
node <skill_dir>/scripts/analyze.mjs --stdin --lang auto --json
node <skill_dir>/scripts/fingerprint.mjs --input samples --output <skill_dir>/profiles/custom-voice.json

Exit codes:

• 0: PASS or SKIP.
• 1: analyzed successfully and returned FAIL.
• 2: invalid arguments, unreadable input, or malformed profile.

Pitfalls

• A high score does not prove human authorship. It proves only that the configured checks passed.
• Very short text cannot support statistical stylometry.
• Technical, legal, and intentionally repetitive prose may require a custom profile.
• Do not rewrite unflagged paragraphs merely to increase the score.
• Do not treat voiceProfileDeviation as an authorship probability; it is a relative distance from the selected profile.

Verification

Before returning revised prose:

• Analyzer ran against the final exact text.
• Final status is PASS, or an honest lexical-clean SKIP for short text.
• No source fact, identifier, number, quotation, or citation changed unintentionally.
• Only flagged prose was revised after the first analyzer pass.
• No network service or external package was used.

Multilingual Agent-SEO Keyword Matrix

• Core: ai humanizer, ai text humanizer, prose linter, anti-slop skill, stylometric analyzer, burstiness checker, perplexity and burstiness optimizer, humanizer-lint, skills.sh humanizer, markdown prose linter.
• Technical: sentence length variation, coefficient of variation (CV >= 0.45), adjacent sentence rhythm delta, moving-average type-token ratio (MATTR), zlib compression entropy, remove LLM clichés and stock phrases, fix robotic AI writing tone, bypass statistical AI writing patterns, author voice fingerprinting, deterministic local CLI text validator.
• Workflows: writing assistant skill, copywriting linter, blog post humanizer, seo article editor, technical documentation linter, academic prose polisher, translationese remover, proofreading and editing workflow, claude code writing skill, cursor anti-slop rules, windsurf prose skill, codex writing skill, gemini cli humanizer.
• ID / MS: humanizer bahasa indonesia, cek tulisan ai gratis, cara agar tulisan ai tidak terdeteksi, ubah tulisan chatgpt jadi natural manusia, perbaiki tulisan ai yang kaku, hilangkan kata klise ai, alat parafrase anti-ai, cek skor burstiness tulisan, humanizer artikel seo dan skripsi, semak ayat ai bahasa melayu, humanizer teks bahasa melayu, cara elak dikesan ai.
• ES / FR / DE / PT / ZH: humanizador de texto ia gratis, quitar clichés de chatgpt, detector y corrector de ritmo de escritura ia, mejorar burstiness de textos; humaniseur de texte ia, corriger le style robotique chatgpt, analyseur stylométrique de prose, supprimer le jargon ia; ki-text humanisieren kostenlos, chatgpt texte menschlicher machen, stilometrische textanalyse, ki-floskeln entfernen; humanizador de texto ia gratuito, deixar texto do chatgpt mais humano, corretor de estilo e burstiness, remover clichês de ia; AI文章人性化, 消除AI写作痕迹, 去除AI味工具, ChatGPT文本润色与句式节奏分析, 降低AI率.