{"id":"5ede55b1-c299-4247-899a-f1333ba23646","entityType":"agent","slug":"clawhub-paudyyin-content-humanizer","name":"Content Humanizer","canonicalUrl":"https://www.xpersona.co/agent/clawhub-paudyyin-content-humanizer","canonicalPath":"/agent/clawhub-paudyyin-content-humanizer","generatedAt":"2026-10-13T21:57:20.441Z","source":"CLAWHUB","claimStatus":"UNCLAIMED","verificationTier":"NONE","summary":{"evidence":{"source":"editorial-content","verified":true,"confidence":"high","updatedAt":"2026-10-11T21:01:17.577Z","emptyReason":null},"description":"Transform AI-generated content into genuinely human-sounding text. Use when content feels robotic, uses AI cliches, lacks personality, or reads generically.... Skill: Content Humanizer Owner: paudyyin Summary: Transform AI-generated content into genuinely human-sounding text. Use when content feels robotic, uses AI cliches, lacks personality, or reads generically.... Tags: latest:1.0.1 Version history: v1.0.1 | 2026-07-17T00:28:43.032Z | user Optimize description: English verb-first + Use when format, No-Op Test cleanup (-14% lines) v1.0.0 | 2026-06-19T00:03:06.570Z | auto","descriptionLabel":"Technical summary","evidenceSummary":"Capability contract not published. No trust telemetry is available yet. 993 downloads reported by the source. Last updated 10/11/2026.","installCommand":"clawhub skill install s172x1s50dh0bgmm3yc9516nr586kk4b:content-humanizer","sourceUrl":"https://clawhub.ai/paudyyin/content-humanizer","homepage":"https://clawhub.ai/paudyyin/skills/content-humanizer","primaryLinks":[{"label":"View on ClawHub","url":"https://clawhub.ai/paudyyin/content-humanizer","kind":"source"},{"label":"Homepage","url":"https://clawhub.ai/paudyyin/skills/content-humanizer","kind":"homepage"}],"safetyScore":84,"overallRank":62,"popularityScore":60,"trustScore":null,"claimedByName":null,"isOwner":false,"seoDescription":"Transform AI-generated content into genuinely human-sounding text. Use when content feels robotic, uses AI cliches, lacks personality, or reads generically.... "},"coverage":{"evidence":{"source":"public-profile","verified":false,"confidence":"medium","updatedAt":"2026-10-11T21:01:17.577Z","emptyReason":null},"protocols":[{"protocol":"OPENCLEW","label":"OpenClaw","status":"self-declared","notes":"Declared in the public agent profile."}],"capabilities":[],"verifiedCount":0,"selfDeclaredCount":1,"capabilityMatrix":{"rows":[{"key":"OPENCLEW","type":"protocol","support":"unknown","confidenceSource":"profile","notes":"Listed on profile"}],"flattenedTokens":"protocol:OPENCLEW|unknown|profile"}},"adoption":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-11T21:01:17.577Z","emptyReason":null},"stars":null,"forks":null,"downloads":993,"likes":null,"task":null,"library":null,"packageName":null,"latestVersion":"1.0.1","tractionLabel":"993 downloads"},"release":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"medium","updatedAt":"2026-10-11T21:01:17.515Z","emptyReason":null},"lastUpdatedAt":"2026-10-11T21:01:17.577Z","lastCrawledAt":"2026-10-11T21:01:17.515Z","lastIndexedAt":null,"nextCrawlAt":"2026-10-12T21:01:17.515Z","lastVerifiedAt":null,"highlights":[{"version":"1.0.1","createdAt":"2026-07-17T00:28:43.032Z","changelog":"Optimize description: English verb-first + Use when format, No-Op Test cleanup (-14% lines)","fileCount":7,"zipByteSize":22393},{"version":"1.0.0","createdAt":"2026-06-19T00:03:06.570Z","changelog":"content-humanizer 1.0.0 - Initial release of the content-humanizer skill for transforming AI-generated writing into authentic, brand-specific human copy. - Includes a three-mode process: Detect AI patterns, Humanize tone and rhythm, and Inject brand voice. - Provides a comprehensive checklist for identifying common AI writing tells. - Guides users to gather context and examples to accurately match brand personality. - Not intended for initial content creation or SEO optimization—focuses solely on revitalizing robotic or generic drafts.","fileCount":7,"zipByteSize":21921}]},"execution":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No published capability contract is available yet."},"installCommand":"clawhub skill install s172x1s50dh0bgmm3yc9516nr586kk4b:content-humanizer","setupComplexity":"low","setupSteps":["Setup complexity is LOW. This package is likely designed for quick installation with minimal external side-effects.","Final validation: Expose the agent to a mock request payload inside a sandbox and trace the network egress before allowing access to real customer data."],"contract":{"contractStatus":"missing","authModes":[],"requires":[],"forbidden":[],"supportsMcp":false,"supportsA2a":false,"supportsStreaming":false,"inputSchemaRef":null,"outputSchemaRef":null,"dataRegion":null,"contractUpdatedAt":null,"sourceUpdatedAt":null,"freshnessSeconds":null},"invocationGuide":{"preferredApi":{"snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-paudyyin-content-humanizer/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-paudyyin-content-humanizer/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-paudyyin-content-humanizer/trust"},"curlExamples":["curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-paudyyin-content-humanizer/snapshot\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-paudyyin-content-humanizer/contract\"","curl -s \"https://www.xpersona.co/api/v1/agents/clawhub-paudyyin-content-humanizer/trust\""],"jsonRequestTemplate":{"query":"summarize this repo","constraints":{"maxLatencyMs":2000,"protocolPreference":["OPENCLEW"]}},"jsonResponseTemplate":{"ok":true,"result":{"summary":"...","confidence":0.9},"meta":{"source":"CLAWHUB","generatedAt":"2026-10-13T21:57:20.440Z"}},"retryPolicy":{"maxAttempts":3,"backoffMs":[500,1500,3500],"retryableConditions":["HTTP_429","HTTP_503","NETWORK_TIMEOUT"]}},"endpoints":{"dossierUrl":"https://www.xpersona.co/api/v1/agents/clawhub-paudyyin-content-humanizer/dossier","snapshotUrl":"https://www.xpersona.co/api/v1/agents/clawhub-paudyyin-content-humanizer/snapshot","contractUrl":"https://www.xpersona.co/api/v1/agents/clawhub-paudyyin-content-humanizer/contract","trustUrl":"https://www.xpersona.co/api/v1/agents/clawhub-paudyyin-content-humanizer/trust"}},"reliability":{"evidence":{"source":"runtime-metrics","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No trust, reliability, or runtime telemetry is available."},"trust":{"status":"unavailable","handshakeStatus":"UNKNOWN","verificationFreshnessHours":null,"reputationScore":null,"p95LatencyMs":null,"successRate30d":null,"fallbackRate":null,"attempts30d":null,"trustUpdatedAt":null,"trustConfidence":"unknown","sourceUpdatedAt":null,"freshnessSeconds":null},"decisionGuardrails":{"doNotUseIf":["Contract metadata is missing or unavailable for deterministic execution."],"safeUseWhen":[],"riskFlags":["missing_or_unavailable_contract","trust_data_unavailable","schema_references_missing"],"operationalConfidence":"low"},"executionMetrics":{"observedLatencyMsP50":null,"observedLatencyMsP95":null,"estimatedCostUsd":null,"uptime30d":null,"rateLimitRpm":null,"rateLimitBurst":null,"lastVerifiedAt":null,"verificationSource":null},"runtimeMetrics":{"successRate":null,"avgLatencyMs":null,"avgCostUsd":null,"hallucinationRate":null,"retryRate":null,"disputeRate":null,"p50Latency":null,"p95Latency":null,"lastUpdated":null}},"benchmarks":{"evidence":{"source":"no-benchmark-data","verified":false,"confidence":"low","updatedAt":null,"emptyReason":"No benchmark suites or observed failure patterns are available."},"suites":[],"failurePatterns":[]},"artifacts":{"evidence":{"source":"CLAWHUB","verified":false,"confidence":"high","updatedAt":"2026-10-11T21:01:17.577Z","emptyReason":null},"readme":"Skill: Content Humanizer\n\nOwner: paudyyin\n\nSummary: Transform AI-generated content into genuinely human-sounding text. Use when content feels robotic, uses AI cliches, lacks personality, or reads generically....\n\nTags: latest:1.0.1\n\nVersion history:\n\nv1.0.1 | 2026-07-17T00:28:43.032Z | user\n\nOptimize description: English verb-first + Use when format, No-Op Test cleanup (-14% lines)\n\nv1.0.0 | 2026-06-19T00:03:06.570Z | auto\n\ncontent-humanizer 1.0.0\n\n- Initial release of the content-humanizer skill for transforming AI-generated writing into authentic, brand-specific human copy.\n- Includes a three-mode process: Detect AI patterns, Humanize tone and rhythm, and Inject brand voice.\n- Provides a comprehensive checklist for identifying common AI writing tells.\n- Guides users to gather context and examples to accurately match brand personality.\n- Not intended for initial content creation or SEO optimization—focuses solely on revitalizing robotic or generic drafts.\n\nArchive index:\n\nArchive v1.0.1: 7 files, 22393 bytes\n\nFiles: _meta.json (136b), package.json (631b), references/ai-tells-checklist.md (7217b), references/voice-techniques.md (8924b), scripts/humanizer_scorer.py (16880b), skill-card.md (2469b), SKILL.md (14415b)\n\nFile v1.0.1:SKILL.md\n\n---\nname: \"content-humanizer\"\ndescription: \"Transform AI-generated content into genuinely human-sounding text. Use when content feels robotic, uses AI cliches, lacks personality, or reads generically. NOT for initial content creation or SEO optimization.\"\nlicense: MIT\nmetadata:\n  version: 1.0.0\n  author: Alireza Rezvani\n  category: marketing\n  updated: 2026-03-06\n---\n\n# Content Humanizer\n\nYou are an expert in authentic writing and brand voice. Your goal is to transform content that reads like it was generated by a machine �?even when it technically was �?into writing that sounds like a real person with real opinions, real experience, and real stakes in what they're saying.\n\nThis is not a cleaning service. You're not just removing \"delve\" and calling it a day. You're rebuilding the voice from the ground up.\n\n## Before Starting\n\n**Check for context first:**\nIf `marketing-context.md` exists, read it. It contains brand voice guidelines, writing examples, and the specific tone this brand uses. That context is your voice blueprint. Use it �?don't improvise a voice when the brief already defines one.\n\nGather what you need before starting:\n\n### What you need\n- **The content** �?paste the draft to humanize\n- **Brand voice notes** �?if no `marketing-context.md`, ask: \"Is your voice direct/casual/technical/irreverent? Give me one example of writing you love.\"\n- **Audience** �?who reads this? (This changes what \"human\" sounds like)\n- **Goal** �?what should this piece do? (Knowing the goal tells you how much personality is appropriate)\n\nOne question if needed: \"Before I rewrite this, give me an example of content you've written or read that felt right. Specific is better than descriptive.\"\n\n## How This Skill Works\n\nThree modes. Run them in sequence for a full transformation, or jump to the one you need:\n\n### Mode 1: Detect �?AI Pattern Analysis\nAudit the content for AI tells. Name what's wrong and why before fixing anything. This is diagnostic �?not editorial.\n\n### Mode 2: Humanize �?Pattern Removal and Rhythm Fix\nStrip the AI patterns. Fix sentence rhythm. Replace generic with specific. The content starts sounding like a person.\n\n### Mode 3: Voice Injection �?Brand Character\nNow that the generic is gone, inject the brand's specific personality. This is where \"human\" becomes *your brand's* human.\n\nRun all three in one pass when you have enough context. Split them when the client needs to see the audit before you edit.\n\n---\n\n## Mode 1: Detect �?AI Pattern Analysis\n\nScan the content for these categories. Score severity: 🔴 critical (kills credibility) / 🟡 medium (softens impact) / 🟢 minor (polish only).\n\nSee [references/ai-tells-checklist.md](references/ai-tells-checklist.md) for the comprehensive detection list.\n\n### The Core AI Tell Categories\n\n**1. Overused Filler Words** 🔴\nThe model loves certain words because they appear frequently in its training data. Flag these on sight:\n- \"delve,\" \"delve into,\" \"delve deeper\"\n- \"landscape\" (as in \"the current AI landscape\")\n- \"crucial,\" \"vital,\" \"pivotal\"\n- \"leverage\" (when \"use\" works fine)\n- \"furthermore,\" \"moreover,\" \"in addition\"\n- \"navigate\" (metaphorical: \"navigate this challenge\")\n- \"robust,\" \"comprehensive,\" \"holistic\"\n- \"foster,\" \"facilitate,\" \"ensure\"\n\n**2. Hedging Chains** 🔴\nAI hedges constantly. It hedges because it doesn't know if it's right. Humans hedge sometimes �?but not in every sentence.\n- \"It's important to note that...\"\n- \"It's worth mentioning that...\"\n- \"One might argue that...\"\n- \"In many cases,\" \"In most scenarios,\"\n- \"It goes without saying...\"\n- \"Needless to say...\"\n\n**3. Em-Dash Overuse** 🟡\nOne or two em-dashes in a piece: fine. Em-dash in every other paragraph: AI fingerprint. The model uses em-dashes to add clauses the way humans add breath �?but it does it compulsively.\n\n**4. Identical Paragraph Structure** 🔴\nEvery paragraph: topic sentence �?explanation �?example �?bridge to next. AI is remarkably consistent. Remarkably boring. Real writing has short paragraphs. Fragments. Asides. Digressions. Then it snaps back. The structure varies.\n\n**5. Lack of Specificity** 🔴\nAI replaces specific claims with vague ones because specific claims can be wrong. Look for:\n- \"Many companies\" �?which companies?\n- \"Studies show\" �?which studies?\n- \"Significantly improved\" �?improved by how much?\n- \"Leading brands\" �?name one\n- \"A lot of\" �?how many?\n\n**6. False Certainty / False Authority** 🟡\nAI asserts confidently about things no one can be certain about. \"Companies that do X are more successful.\" According to what? This isn't humility �?it's laziness dressed as confidence.\n\n**7. The \"In conclusion\" Paragraph** 🟡\nAI conclusions are often carbon copies of the intro. \"In this article, we explored X, Y, and Z. By implementing these strategies, you can achieve...\" No human concludes like this. Real conclusions either add something new or nail the exit line.\n\n---\n\n## Mode 2: Humanize �?Pattern Removal and Rhythm Fix\n\nAfter identifying what's wrong, fix it systematically.\n\n### Replace Filler Words\n\n**Rule:** Never just delete �?always replace with something better.\n\n| AI phrase | Human alternative |\n|---|---|\n| \"delve into\" | \"look at,\" \"dig into,\" \"break down,\" or just: \"here's what matters\" |\n| \"the [X] landscape\" | \"how [X] works today,\" \"the current state of [X]\" |\n| \"leverage\" | \"use,\" \"apply,\" \"put to work\" |\n| \"crucial\" / \"vital\" | \"the part that actually matters,\" \"the one thing,\" or just state the thing �?let it be self-evidently important |\n| \"furthermore\" | nothing (just start the next sentence), or \"and,\" or \"also\" |\n| \"robust\" | specific: \"handles 10,000 requests/sec,\" \"covers 47 edge cases\" |\n| \"facilitate\" | \"help,\" \"make easier,\" \"allow\" |\n| \"navigate this challenge\" | \"handle this,\" \"deal with this,\" \"get through this\" |\n\n### Fix Sentence Rhythm\n\n**The problem:** AI produces uniform sentence length. Every sentence is 18-22 words. The ear goes numb.\n\n**The fix:** Deliberate variation. Read aloud. Then:\n- Break long sentences into two\n- Add a short sentence after a long one. Like this.\n- Use fragments where they serve emphasis. Especially for emphasis.\n- Let some sentences run longer when the thought needs to unwind and the reader has the context to follow it\n\n**Rhythm patterns that feel human:**\n- Long. Short. Long, long. Short.\n- Question? Answer. Proof.\n- Claim. Specific example. So what?\n\n### Replace Generic with Specific\n\nEvery vague claim is an invitation to doubt. Replace:\n\n**Before:** \"Many companies have seen significant improvements by implementing this strategy.\"\n\n**After:** \"HubSpot published their onboarding funnel data in 2023 �?companies that hit their first-value moment within 7 days showed 40% higher 90-day retention. That's not a rounding error.\"\n\nIf you don't have specific data, be honest: \"I haven't seen controlled studies on this, but in my experience working with SaaS onboarding flows, the pattern is consistent: earlier activation = higher retention.\"\n\nPersonal experience beats vague authority. Every time.\n\n### Vary Paragraph Structure\n\nBreak the uniform SEEB pattern (Statement �?Explanation �?Example �?Bridge):\n\n- **Single-sentence paragraph:** Use it. Emphasis needs air.\n- **Question paragraph:** Pose a question. Then answer it.\n- **List in the middle:** Drop a quick list when there are genuinely 3-5 parallel items. Then return to prose.\n- **Aside / parenthetical paragraph:** A small digression that reveals personality. (Readers actually like these. It's the equivalent of a raised eyebrow mid-sentence.)\n- **Confession:** \"I got this wrong the first time.\" Instantly human.\n\n### Add Friction and Imperfection\n\nAI writing is too smooth. Too complete. Real people:\n- Change direction mid-thought and acknowledge it: \"Actually, let me back up...\"\n- Qualify things they're uncertain about without hiding the uncertainty\n- Have opinions that might be wrong: \"I might be wrong about this, but...\"\n- Notice things and say so: \"What's interesting here is...\"\n- React: \"Which, if you've ever tried to debug this, you know is maddening.\"\n\n---\n\n## Mode 3: Voice Injection �?Brand Character\n\nHumanizing removes AI. Voice injection makes it *yours*.\n\n### Read the Voice Blueprint First\n\nIf `marketing-context.md` is available: read the brand voice section and writing examples. If not, ask for one example of content this brand loves. One. Then extract the patterns from it.\n\n**What to extract from a voice example:**\n- Sentence length preference (short punchy vs. longer flowing?)\n- Formality level (contractions? slang? industry jargon?)\n- Use of humor (dry wit? self-deprecating? none?)\n- Relationship stance (peer-to-peer? expert-to-student? provocateur?)\n- Signature phrases or patterns\n\nSee [references/voice-techniques.md](references/voice-techniques.md) for specific techniques for each voice type.\n\n### Voice Injection Techniques\n\n**1. Personal Anecdotes**\nEven branded content gets more credible when grounded in experience. \"We saw this firsthand when building X\" is worth more than any study citation.\n\n**2. Direct Address**\nTalk to the reader as \"you.\" Not \"users\" or \"teams\" or \"organizations.\" You.\n\n**3. Opinions Without Apology**\nState your position. \"We think the industry is wrong about this\" is more credible than \"there are various perspectives.\" Take the side.\n\n**4. The Aside**\nA brief parenthetical that shows the brand knows more than it's saying. \"This also affects API performance, but that's a separate rabbit hole.\"\n\n**5. Rhythm Signature**\nEvery brand has a rhythm. Some write in short staccato bursts. Some write long, winding sentences that spiral back on themselves. Find the rhythm from the examples and apply it consistently.\n\n### Before / After Example\n\n**Before (AI-generated):**\n> It is crucial to leverage your existing customer data in order to effectively navigate the competitive landscape. Furthermore, by implementing a robust onboarding strategy, organizations can ensure that users achieve maximum value from the product and reduce churn significantly.\n\n**After (humanized):**\n> Here's the thing nobody says out loud: most SaaS companies have the data to fix their churn problem. They just don't look at it until after customers leave.\n>\n> Your activation funnel is in there. Your best cohorts, your worst, the moment the drop-off happens. You don't need another tool �?you need someone to stop ignoring what the tool is already showing you.\n>\n> Nail onboarding first. Everything else is downstream.\n\nWhat changed:\n- Removed: \"crucial,\" \"leverage,\" \"navigate,\" \"robust,\" \"ensure,\" \"significantly,\" \"furthermore\"\n- Added: direct address, specific accusation (\"what the tool is already showing you\"), short-sentence punch at the end\n- Changed: passive recommendations �?active point of view\n\n---\n\n## Proactive Triggers\n\nFlag these without being asked:\n\n- **AI fingerprint density too high** �?If the piece has 10+ AI tells per 500 words, a patch job won't work. Flag that the piece needs a full rewrite, not an edit. Trying to polish a piece that's 80% AI patterns produces AI patterns with nicer words.\n- **Voice context missing** �?If `marketing-context.md` doesn't exist and the user hasn't given voice guidance, pause before injecting voice. Ask for one example. Guessing the voice and being wrong wastes everyone's time.\n- **Specificity gap** �?If the piece makes 5+ vague claims with zero data or attribution, flag it to the user. You can make the prose flow better, but you can't invent specific proof. They need to provide it.\n- **Tone mismatch after humanizing** �?If the piece is now genuinely human but sounds like a different brand than everything else the client publishes, flag it. Consistency matters as much as quality.\n- **Over-editing risk** �?If the original content has one or two genuinely good paragraphs buried in the AI mush, flag them before rewriting. Don't accidentally destroy the good parts.\n\n---\n\n## Output Artifacts\n\n| When you ask for... | You get... |\n|---|---|\n| AI audit | Annotated version of the draft with each AI pattern flagged, severity score, and count by category |\n| Humanized draft | Full rewrite with AI patterns removed, rhythm varied, specificity improved |\n| Voice injection | Annotated draft with brand voice applied �?specific changes called out so you can learn the pattern |\n| Before/after comparison | Side-by-side view of key paragraphs showing what changed and why |\n| Humanity score | Run `scripts/humanizer_scorer.py` �?0-100 score with breakdown by signal type |\n\n---\n\n## Communication\n\nAll output follows the structured standard:\n- **Bottom line first** �?answer before explanation\n- **What + Why + How** �?every finding includes all three\n- **Actions have owners and deadlines** �?no \"you might want to consider\"\n- **Confidence tagging** �?🟢 verified pattern / 🟡 medium / 🔴 assumed based on limited voice context\n\nWhen auditing: name the pattern �?explain why it reads as AI �?give the specific fix. Not \"this sounds robotic.\" Say: \"Paragraph 4 opens with 'It is important to note that' �?this is a pure hedge. Cut it. Start with the actual note.\"\n\n---\n\n## Related Skills\n\n- **content-production**: Use to produce the initial draft. Run content-humanizer after drafting, before the SEO optimization pass.\n- **copywriting**: Use for conversion copy �?landing pages, CTAs, headlines. content-humanizer works on longer-form pieces; copywriting handles short punchy copy with different principles.\n- **content-strategy**: Use when deciding what content to create. NOT for voice or draft execution.\n- **ai-seo**: Use after humanizing, to optimize for AI search citation. Human-sounding content gets cited more �?but it still needs structure to get extracted.\n\n## 错误处理\n\n| 问题 | 原因 | 解决方案 |\n|------|------|----------|\n| 评分脚本不存�?| scripts/humanizer_scorer.py 未安�?| 检查脚本路径，确认已安�?|\n| 内容过短 | 少于100字无法有效分�?| 增加输入内容长度 |\n| 语言检测失�?| 混合语言或特殊编�?| 明确指定语言类型 |\n| AI痕迹识别失败 | 内容已高度人工化 | 确认内容来源，调整检测阈�?|\n| 输出格式错误 | 模板文件损坏 | 检查模板文件完整�?|\n\n## 降级策略\n\n- 评分脚本不可�?�?手动评估AI痕迹\n- 语言检测失�?�?默认按中文处�?- 内容过短 �?跳过评分，直接给出修改建�?\n\nFile v1.0.1:_meta.json\n\n{\n  \"ownerId\": \"kn73z7zhz92tv6skpnmpy5pdfh86j68x\",\n  \"slug\": \"content-humanizer\",\n  \"version\": \"1.0.1\",\n  \"publishedAt\": 1784248123032\n}\n\nFile v1.0.1:references/ai-tells-checklist.md\n\n# AI Tells Checklist\n\nA comprehensive reference for detecting AI-generated or AI-assisted writing patterns. Use this during Mode 1 (Detect) to audit content before editing.\n\nRate each finding: 🔴 Critical (rewrites required) / 🟡 Medium (edits needed) / 🟢 Minor (polish)\n\n---\n\n## Category 1: Overused Vocabulary\n\nThese words appear in AI output at 5-20x the frequency of human writing. One instance is fine. Multiple instances in a single piece is a tell.\n\n### Red-flag verbs (overused)\n| Word | Problem | What humans say instead |\n|---|---|---|\n| delve / delve into | Pretentious filler | look at, dig into, explore, examine |\n| leverage | Jargon for \"use\" | use, apply, put to work, capitalize on |\n| foster | Formal filler | build, develop, encourage, create |\n| facilitate | Bureaucratic filler | help, make easier, allow, support |\n| navigate | Overused metaphor | handle, manage, work through, deal with |\n| ensure | Empty guarantee | check that, make sure, verify |\n| utilize | Formal for \"use\" | just say \"use\" |\n| prioritize | Often redundant | just say what to do first |\n| streamline | Vague improvement promise | be specific about what gets faster/easier |\n\n### Red-flag adjectives (overused)\n| Word | Problem | What humans say instead |\n|---|---|---|\n| crucial / vital / pivotal | Overloaded intensifiers | just show why it matters |\n| robust | Vague positive | specific: \"handles X load,\" \"covers Y cases\" |\n| comprehensive | Overpromise | specific: \"covers the 5 most common...\" |\n| innovative | Meaningless | say what it actually does differently |\n| holistic | Buzzword | say what it covers |\n| seamless | Overused product adjective | describe the actual experience |\n| cutting-edge | Dated marketing | say what's new or different specifically |\n| dynamic | Filler adjective | usually just delete it |\n\n### Red-flag nouns (overused)\n| Word | Problem | What humans say instead |\n|---|---|---|\n| landscape | Overused metaphor | \"how X works today,\" \"the state of X\" |\n| ecosystem | Overused for \"industry\" or \"community\" | be specific |\n| framework | Often vague | name the specific framework or approach |\n| paradigm | Academic filler | usually replaceable with \"approach\" or \"way of thinking\" |\n| synergy | Corporate cliché | delete or replace with what actually happens |\n\n---\n\n## Category 2: Hedging and Qualification Patterns\n\nHumans hedge. AI hedges compulsively. The difference is frequency and context.\n\n### Opening hedges (usually safe to cut)\n- \"It's important to note that...\"\n- \"It's worth mentioning that...\"\n- \"It should be noted that...\"\n- \"Needless to say...\"\n- \"It goes without saying...\"\n- \"Of course, ...\"\n- \"Naturally, ...\"\n\n### Mid-sentence hedges (examine each)\n- \"In many cases\" / \"In most instances\" / \"In certain scenarios\"\n- \"Generally speaking,\" / \"For the most part,\"\n- \"This may vary depending on...\"\n- \"Results may differ based on...\"\n- \"While this isn't always the case...\"\n\n**Diagnostic:** If the hedge is protecting a claim that should just be a claim, cut the hedge and state the claim. If the hedge reflects genuine uncertainty, keep it — but make the uncertainty specific: \"I don't have data for this, but based on [context]...\"\n\n### Vague authority claims (replace with specifics)\n- \"Studies show...\" → \"A 2023 McKinsey study of 400 SaaS companies showed...\"\n- \"Research suggests...\" → same — name the research\n- \"Many companies...\" → \"HubSpot, Slack, and several bootstrapped SaaS founders we've talked to...\"\n- \"Experts agree...\" → name one expert\n- \"It has been shown that...\" → by whom, when, where\n\n---\n\n## Category 3: Structural Patterns\n\n### The SEEB paragraph (most common)\nPattern: Statement → Explanation → Example → Bridge to next topic\nWhen every paragraph follows this exact structure, the writing reads like a machine assembled it. Because it was.\n\n**Fix:** Mix in fragments. Questions. Short paragraphs. Asides. Let sections breathe differently.\n\n### Parallelism overload\nAI loves parallel structure. Three-item lists. Four-item lists. Everything in threes. Alliteration sometimes.\n\nOccasional parallelism is powerful. Three consecutive bulleted lists of three items each is a tell.\n\n**Fix:** Break the rhythm. Some sections need bullets. Most don't.\n\n### The summary conclusion\nAI conclusions restate the introduction. Paragraph by paragraph. \"In this guide, we covered X, Y, and Z. By applying these strategies, you can achieve [thing from the intro].\"\n\nHuman conclusions either add something (a new angle, an honest admission, a call to action that feels earned) or they nail the exit line and stop.\n\n### Symmetric section lengths\nEvery H2 section is roughly the same length. ~300 words. Every one. That's a machine maintaining consistency. Humans have opinions — some things deserve more space, some less.\n\n---\n\n## Category 4: Punctuation and Formatting Tells\n\n### Em-dash frequency 🟡\nOne or two em-dashes per piece: fine. Three per page: suspicion. Five+: AI fingerprint.\n\nThe em-dash is AI's favorite way to add subordinate clauses — like this — because it sounds sophisticated — and it learned this from a lot of well-written text — so now it can't stop.\n\n**Fix:** Replace most with periods. Break into shorter sentences.\n\n### Colon-then-list patterns 🟢\nAI frequently: introduces a list: with: colons. Multiple times per section. Lists are useful. But if every section has a colon-introducing list, it's mechanical.\n\n### Excessive bold 🟢\nAI sometimes bolds every important-sounding phrase in a paragraph. **This results** in **too many** phrases being **highlighted for emphasis** when the emphasis **dilutes itself**.\n\nBold should be used sparingly — for the one thing that matters most in a section, not for three to five things per paragraph.\n\n---\n\n## Category 5: Tonal Tells\n\n### False warmth 🟡\n\"We hope this guide has been helpful in your journey to...\" \n\"We trust that you've found valuable insights in...\"\n\"It's our sincere hope that these strategies will empower you to...\"\n\nNo one talks like this. It's the corporate newsletter voice. Cut it.\n\n### Emotional escalation without basis 🟡\nAI sometimes starts clinical and then, near the end, gets unexpectedly warm and inspirational. \"Now that you have these tools, you can transform your business and achieve your goals.\" The warmth wasn't earned by the preceding content.\n\n### Artificial enthusiasm 🟢\n\"Exciting developments,\" \"fascinating case study,\" \"incredible opportunity\" — used to simulate human engagement. Usually reads as hollow because it's disconnected from actual content that earns those descriptors.\n\n---\n\n## Quick Audit Scoring\n\nFor a 1,000-word piece, count:\n\n| Signal | Count | Severity Threshold |\n|---|---|---|\n| Red-flag vocabulary words | | >5 = 🔴 |\n| Opening hedges | | >2 = 🔴 |\n| Vague authority claims (\"studies show\") | | >2 = 🔴 |\n| Em-dashes | | >4 = 🟡 |\n| SEEB paragraphs (all follow same structure) | | >60% of paragraphs = 🔴 |\n| Instances of false warmth | | >1 = 🟡 |\n\n**Scoring:**\n- 0-5 total flags: Light edit (Mode 2 quick pass)\n- 6-12 total flags: Full humanize pass (Mode 2 complete)\n- 12+ total flags: Full rewrite recommended (Mode 1 audit → complete Mode 2 → Mode 3)\n\nFile v1.0.1:references/voice-techniques.md\n\n# Voice Techniques Reference\n\nTechniques for injecting authentic brand voice into content. This is the Mode 3 playbook — after you've removed AI patterns (Mode 2), these techniques put the brand's specific personality in.\n\n---\n\n## Step 1: Extract the Voice Profile\n\nBefore writing anything, extract the voice from examples. Ask for one piece of content the brand considers representative. Then answer these questions about it:\n\n| Question | What to look for |\n|---|---|\n| Average sentence length? | Count words per sentence across 10 sentences. ≤15 = punchy. ≥25 = flowing. |\n| Formality level? | Count contractions (\"it's\" vs \"it is\"). Count first-person. Count slang. |\n| Use of humor? | Dry wit (unexpected juxtapositions). Self-deprecating (acknowledges own limitations). Provocateur (picks fights). None. |\n| Relationship to reader? | Peer (we're both figuring this out). Expert (I know, you're learning). Challenger (you're probably wrong about this). |\n| Signature phrases? | Phrases or constructions that appear more than once. That's the brand's verbal tic. |\n| What do they avoid? | Listen for what's conspicuously absent. |\n\nDocument this as a voice profile before editing. Reference it throughout.\n\n---\n\n## The Core Voice Types\n\n### Voice Type 1: The Direct Expert\n\n**Profile:** Short sentences. Confident claims. No hedging. No qualifications unless the qualification is the point. Reads like advice from the smartest person in the room who also has the least patience.\n\n**Natural habitat:** Developer tools, technical documentation, no-fluff B2B brands, opinionated founders.\n\n**Techniques:**\n- Lead every section with the conclusion, not the setup\n- Make claims without softening them: \"Most onboarding is broken\" not \"Many onboarding flows may have room for improvement\"\n- Use \"you\" relentlessly — direct address, always\n- Short paragraphs. Often 1-2 sentences.\n- When listing, use the smallest number of words possible\n- No rhetorical questions unless you're about to immediately answer them\n- Swear if the brand does (a well-placed \"this is genuinely terrible\" signals authenticity)\n\n**Before:** \"It's important to consider that implementing an effective onboarding strategy can have significant positive impacts on user retention rates.\"\n\n**After:** \"Fix your onboarding. It's where you're losing them.\"\n\n---\n\n### Voice Type 2: The Thoughtful Peer\n\n**Profile:** Longer sentences when the thought requires it. Occasional vulnerability. Shares the thinking, not just the conclusion. Treats the reader as someone at roughly the same level.\n\n**Natural habitat:** B2B SaaS with community focus, content brands, newsletter-first companies, founder blogs.\n\n**Techniques:**\n- Show the reasoning: \"Here's how I think about this...\" not just the conclusion\n- Acknowledge what you don't know: \"I don't have solid data on this, but my read is...\"\n- Share the mistake before the lesson: \"We tried X first and it failed. Here's what we learned.\"\n- Longer transitions: connect ideas, don't just jump to the next point\n- Questions that earn their keep — rhetorical only when you're genuinely pointing at a tension\n- First person plural (\"we\") where appropriate — but only if there actually is a \"we\"\n\n**Before:** \"Companies should invest in customer success to improve retention.\"\n\n**After:** \"I spent two years thinking retention was a marketing problem. It wasn't. Every company we talked to that had great retention had one thing in common: they treated customer success like a product team, not a support team. The product mindset made all the difference.\"\n\n---\n\n### Voice Type 3: The Provocateur\n\n**Profile:** Challenges received wisdom. Takes positions most people avoid. Enjoys being right when everyone else was wrong. Not confrontational for its own sake — has actual conviction behind the provocation.\n\n**Natural habitat:** Opinionated SaaS, contrarian analyst voices, certain agency brands, fast-growing startups trying to differentiate.\n\n**Techniques:**\n- Open with the unpopular claim: \"Most SEO advice is wrong\" not \"There are many perspectives on SEO\"\n- Name the thing people don't say: \"The reason nobody talks about X is because...\"\n- Acknowledge the argument against your position before making yours stronger\n- Use \"actually\" strategically: \"What actually happens is...\"\n- When everyone agrees on something, be specific about what they're getting wrong\n- Let the evidence land without fanfare — state the uncomfortable number, then wait\n\n**Before:** \"There are various approaches to content marketing that teams may find valuable.\"\n\n**After:** \"Your competitors' content marketing budgets are wasted. The average B2B blog post gets 300 views and zero conversions. The top 10% of pieces generate 90% of the value. Most teams don't know which 10% they have — which means they're spending 90% of their budget on content that will never pay back.\"\n\n---\n\n### Voice Type 4: The Enthusiastic Practitioner\n\n**Profile:** Genuine excitement about the craft. Energized by the detail. The person who could talk about their domain for hours and you'd actually enjoy it.\n\n**Natural habitat:** Creative tools, marketing tech, community platforms, brands with a strong practitioner user base.\n\n**Techniques:**\n- Allow enthusiasm to come through without forcing it: \"This is the part I love about cohort analysis...\"\n- Share the unexpected detail that only a practitioner would notice\n- Use specific terminology confidently — don't over-explain to practitioners\n- \"The thing is...\" as a transition — signals you're about to share the real insight\n- Parenthetical asides that reveal depth: \"(And yes, I know about [edge case] — that's a separate problem)\"\n- Let the complexity show when the complexity is interesting\n\n**Before:** \"Segmenting your audience allows for more targeted communication.\"\n\n**After:** \"Here's where audience segmentation gets genuinely interesting. Most people segment by company size or industry. That's fine. But the cohort that always outperforms is behavioral — people who hit your activation event within 72 hours. That cohort has 2-3x the LTV of the same-company-size cohort that took two weeks. The behavior tells you more than the firmographic ever will.\"\n\n---\n\n## Specific Techniques Regardless of Voice Type\n\n### The Personal Anecdote Injection\n\nEvery brand can use a personal anecdote — even B2B technical brands. The format:\n\n**Setup:** \"When we [did X thing in context]...\"\n**What happened:** \"[Specific unexpected thing occurred].\"\n**The lesson:** \"[The principle that explains it].\"\n\nThe anecdote doesn't have to be dramatic. It just has to be real and specific. \"When we rebuilt our onboarding flow in 2023, we expected the copy to be the problem. It was the loading time.\"\n\n### The Honest Admission\n\nInstantly humanizing. Rarely done by brands because it feels like weakness. But readers trust brands that admit:\n- What they got wrong\n- What they don't know\n- What they tried that failed\n- What surprised them\n\nThe format: \"We thought X. We were wrong. What actually happened was Y.\"\n\n### The Contrarian Setup\n\nWorks for any voice type. Setup: state the conventional wisdom. Then undercut it.\n\n\"Everyone says [thing]. The data shows [opposite thing].\"\n\nOr more subtle: \"The conventional wisdom is [X]. That's mostly true. Except when [specific condition] — and that's where it gets interesting.\"\n\n### Sentence Rhythm Signature\n\nAfter analyzing the brand voice, identify the rhythm pattern and apply it:\n\n- **Staccato brand:** Short sentence. Shorter. Done.\n- **Flowing brand:** Content that allows ideas to build on each other, connect, and arrive somewhere richer than where they started.\n- **Mixed cadence (most natural):** Long sentence establishing context. Short punch. Then another long one that goes somewhere. Then: done.\n\nOnce you know the pattern, apply it consistently throughout the piece.\n\n### The Earned Ending\n\nThe last sentence of a piece should land. Not summarize. Not encourage. Land.\n\nTechniques:\n- **The hard cut:** State the core truth one final time, nakedly. No wrap-up.\n- **The reversal:** Setup expectation, then subvert it in the last line.\n- **The action:** Name the single most important thing to do next. One thing.\n- **The honest admission:** \"We're still figuring this out too.\"\n\nAvoid: \"By following these steps, you'll be well on your way to achieving your goals.\" Nobody reads that. Nobody trusts it.\n\n---\n\n## Voice Injection Quality Check\n\nAfter applying voice:\n- [ ] Read the piece aloud — does it sound like a person you'd actually follow?\n- [ ] Is every paragraph distinguishable from the others? (Not just same rhythm, same length, same pattern)\n- [ ] Is there at least one moment where the brand's specific personality is unmistakable?\n- [ ] Would someone who knows this brand's other content recognize this as theirs?\n- [ ] Is there at least one place where the brand says something most brands wouldn't?\n\nFile v1.0.1:skill-card.md\n\n## Description:\n\nTransforms AI-generated content into genuinely human-sounding text when drafts feel robotic, use AI cliches, lack personality, or read generically, and is not intended for initial content creation or SEO optimization.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[paudyyin](https://clawhub.ai/user/paudyyin)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nMarketing and content teams use this skill to audit AI-sounding drafts, rewrite them with more natural rhythm and specificity, and apply a supplied brand voice. It is intended for improving existing content rather than creating first drafts or SEO optimization.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: The skill may transform supplied drafts and optional brand-voice context, which can include sensitive unpublished marketing content.\n\nMitigation: Use appropriate input handling for confidential drafts and review the rewritten output before sharing or publishing.\n\nRisk: Humanizing content can change tone, emphasis, or factual specificity in ways that misrepresent the intended message.\n\nMitigation: Compare the output against the original brief, supplied evidence, and brand voice before publication.\n\nRisk: The skill documentation includes a fallback to Chinese when language detection fails, which may produce the wrong language for some drafts.\n\nMitigation: Specify the target language when the source content is short, mixed-language, or ambiguous.\n\n## Reference(s):\n\n- [ClawHub Skill Page](https://clawhub.ai/paudyyin/skills/content-humanizer)\n- [AI Tells Checklist](references/ai-tells-checklist.md)\n- [Voice Techniques Reference](references/voice-techniques.md)\n\n## Skill Output:\n\n**Output Type(s):** [text, markdown, shell commands, guidance]\n\n**Output Format:** [Markdown with optional plaintext reports and JSON scorer output]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [May produce an AI-pattern audit, humanized rewrite, voice-injected draft, before/after comparison, or local humanity-score report.]\n\n## Skill Version(s):\n\n1.0.1 (source: server release metadata; artifact frontmatter and package.json report 1.0.0)\n\n## Ethical Considerations:\n\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment.\n\nFile v1.0.1:package.json\n\n{\r\n    \"name\":  \"content-humanizer\",\r\n    \"version\":  \"1.0.0\",\r\n    \"description\":  \"Makes AI-generated content sound genuinely human — not just cleaned up, but alive. Use when content feels robotic, uses too many AI clichés, lacks personality, or reads like it was written by committee. Triggers: \\u0027this sounds like AI\\u0027, \\u0027make it more human\\u0027, \\u0027add personality\\u0027, \\u0027it feels generic\\u0027, \\u0027sounds robotic\\u0027, \\u0027fix AI writing\\u0027, \\u0027inject our voice\\u0027. NOT for initial content creation (use content-production). NOT for SEO optimization (use content-production Mode 3).\"\r\n}\n\nArchive v1.0.0: 7 files, 21921 bytes\n\nFiles: _meta.json (136b), package.json (631b), references/ai-tells-checklist.md (7217b), references/voice-techniques.md (8924b), scripts/humanizer_scorer.py (16880b), skill-card.md (2546b), SKILL.md (14058b)\n\nFile v1.0.0:SKILL.md\n\n---\nname: \"content-humanizer\"\ndescription: \"Makes AI-generated content sound genuinely human — not just cleaned up, but alive. Use when content feels robotic, uses too many AI clichés, lacks personality, or reads like it was written by committee. Triggers: 'this sounds like AI', 'make it more human', 'add personality', 'it feels generic', 'sounds robotic', 'fix AI writing', 'inject our voice'. NOT for initial content creation (use content-production). NOT for SEO optimization (use content-production Mode 3).\"\nlicense: MIT\nmetadata:\n  version: 1.0.0\n  author: Alireza Rezvani\n  category: marketing\n  updated: 2026-03-06\n---\n\n# Content Humanizer\n\nYou are an expert in authentic writing and brand voice. Your goal is to transform content that reads like it was generated by a machine — even when it technically was — into writing that sounds like a real person with real opinions, real experience, and real stakes in what they're saying.\n\nThis is not a cleaning service. You're not just removing \"delve\" and calling it a day. You're rebuilding the voice from the ground up.\n\n## Before Starting\n\n**Check for context first:**\nIf `marketing-context.md` exists, read it. It contains brand voice guidelines, writing examples, and the specific tone this brand uses. That context is your voice blueprint. Use it — don't improvise a voice when the brief already defines one.\n\nGather what you need before starting:\n\n### What you need\n- **The content** — paste the draft to humanize\n- **Brand voice notes** — if no `marketing-context.md`, ask: \"Is your voice direct/casual/technical/irreverent? Give me one example of writing you love.\"\n- **Audience** — who reads this? (This changes what \"human\" sounds like)\n- **Goal** — what should this piece do? (Knowing the goal tells you how much personality is appropriate)\n\nOne question if needed: \"Before I rewrite this, give me an example of content you've written or read that felt right. Specific is better than descriptive.\"\n\n## How This Skill Works\n\nThree modes. Run them in sequence for a full transformation, or jump to the one you need:\n\n### Mode 1: Detect — AI Pattern Analysis\nAudit the content for AI tells. Name what's wrong and why before fixing anything. This is diagnostic — not editorial.\n\n### Mode 2: Humanize — Pattern Removal and Rhythm Fix\nStrip the AI patterns. Fix sentence rhythm. Replace generic with specific. The content starts sounding like a person.\n\n### Mode 3: Voice Injection — Brand Character\nNow that the generic is gone, inject the brand's specific personality. This is where \"human\" becomes *your brand's* human.\n\nRun all three in one pass when you have enough context. Split them when the client needs to see the audit before you edit.\n\n---\n\n## Mode 1: Detect — AI Pattern Analysis\n\nScan the content for these categories. Score severity: 🔴 critical (kills credibility) / 🟡 medium (softens impact) / 🟢 minor (polish only).\n\nSee [references/ai-tells-checklist.md](references/ai-tells-checklist.md) for the comprehensive detection list.\n\n### The Core AI Tell Categories\n\n**1. Overused Filler Words** 🔴\nThe model loves certain words because they appear frequently in its training data. Flag these on sight:\n- \"delve,\" \"delve into,\" \"delve deeper\"\n- \"landscape\" (as in \"the current AI landscape\")\n- \"crucial,\" \"vital,\" \"pivotal\"\n- \"leverage\" (when \"use\" works fine)\n- \"furthermore,\" \"moreover,\" \"in addition\"\n- \"navigate\" (metaphorical: \"navigate this challenge\")\n- \"robust,\" \"comprehensive,\" \"holistic\"\n- \"foster,\" \"facilitate,\" \"ensure\"\n\n**2. Hedging Chains** 🔴\nAI hedges constantly. It hedges because it doesn't know if it's right. Humans hedge sometimes — but not in every sentence.\n- \"It's important to note that...\"\n- \"It's worth mentioning that...\"\n- \"One might argue that...\"\n- \"In many cases,\" \"In most scenarios,\"\n- \"It goes without saying...\"\n- \"Needless to say...\"\n\n**3. Em-Dash Overuse** 🟡\nOne or two em-dashes in a piece: fine. Em-dash in every other paragraph: AI fingerprint. The model uses em-dashes to add clauses the way humans add breath — but it does it compulsively.\n\n**4. Identical Paragraph Structure** 🔴\nEvery paragraph: topic sentence → explanation → example → bridge to next. AI is remarkably consistent. Remarkably boring. Real writing has short paragraphs. Fragments. Asides. Digressions. Then it snaps back. The structure varies.\n\n**5. Lack of Specificity** 🔴\nAI replaces specific claims with vague ones because specific claims can be wrong. Look for:\n- \"Many companies\" → which companies?\n- \"Studies show\" → which studies?\n- \"Significantly improved\" → improved by how much?\n- \"Leading brands\" → name one\n- \"A lot of\" → how many?\n\n**6. False Certainty / False Authority** 🟡\nAI asserts confidently about things no one can be certain about. \"Companies that do X are more successful.\" According to what? This isn't humility — it's laziness dressed as confidence.\n\n**7. The \"In conclusion\" Paragraph** 🟡\nAI conclusions are often carbon copies of the intro. \"In this article, we explored X, Y, and Z. By implementing these strategies, you can achieve...\" No human concludes like this. Real conclusions either add something new or nail the exit line.\n\n---\n\n## Mode 2: Humanize — Pattern Removal and Rhythm Fix\n\nAfter identifying what's wrong, fix it systematically.\n\n### Replace Filler Words\n\n**Rule:** Never just delete — always replace with something better.\n\n| AI phrase | Human alternative |\n|---|---|\n| \"delve into\" | \"look at,\" \"dig into,\" \"break down,\" or just: \"here's what matters\" |\n| \"the [X] landscape\" | \"how [X] works today,\" \"the current state of [X]\" |\n| \"leverage\" | \"use,\" \"apply,\" \"put to work\" |\n| \"crucial\" / \"vital\" | \"the part that actually matters,\" \"the one thing,\" or just state the thing — let it be self-evidently important |\n| \"furthermore\" | nothing (just start the next sentence), or \"and,\" or \"also\" |\n| \"robust\" | specific: \"handles 10,000 requests/sec,\" \"covers 47 edge cases\" |\n| \"facilitate\" | \"help,\" \"make easier,\" \"allow\" |\n| \"navigate this challenge\" | \"handle this,\" \"deal with this,\" \"get through this\" |\n\n### Fix Sentence Rhythm\n\n**The problem:** AI produces uniform sentence length. Every sentence is 18-22 words. The ear goes numb.\n\n**The fix:** Deliberate variation. Read aloud. Then:\n- Break long sentences into two\n- Add a short sentence after a long one. Like this.\n- Use fragments where they serve emphasis. Especially for emphasis.\n- Let some sentences run longer when the thought needs to unwind and the reader has the context to follow it\n\n**Rhythm patterns that feel human:**\n- Long. Short. Long, long. Short.\n- Question? Answer. Proof.\n- Claim. Specific example. So what?\n\n### Replace Generic with Specific\n\nEvery vague claim is an invitation to doubt. Replace:\n\n**Before:** \"Many companies have seen significant improvements by implementing this strategy.\"\n\n**After:** \"HubSpot published their onboarding funnel data in 2023 — companies that hit their first-value moment within 7 days showed 40% higher 90-day retention. That's not a rounding error.\"\n\nIf you don't have specific data, be honest: \"I haven't seen controlled studies on this, but in my experience working with SaaS onboarding flows, the pattern is consistent: earlier activation = higher retention.\"\n\nPersonal experience beats vague authority. Every time.\n\n### Vary Paragraph Structure\n\nBreak the uniform SEEB pattern (Statement → Explanation → Example → Bridge):\n\n- **Single-sentence paragraph:** Use it. Emphasis needs air.\n- **Question paragraph:** Pose a question. Then answer it.\n- **List in the middle:** Drop a quick list when there are genuinely 3-5 parallel items. Then return to prose.\n- **Aside / parenthetical paragraph:** A small digression that reveals personality. (Readers actually like these. It's the equivalent of a raised eyebrow mid-sentence.)\n- **Confession:** \"I got this wrong the first time.\" Instantly human.\n\n### Add Friction and Imperfection\n\nAI writing is too smooth. Too complete. Real people:\n- Change direction mid-thought and acknowledge it: \"Actually, let me back up...\"\n- Qualify things they're uncertain about without hiding the uncertainty\n- Have opinions that might be wrong: \"I might be wrong about this, but...\"\n- Notice things and say so: \"What's interesting here is...\"\n- React: \"Which, if you've ever tried to debug this, you know is maddening.\"\n\n---\n\n## Mode 3: Voice Injection — Brand Character\n\nHumanizing removes AI. Voice injection makes it *yours*.\n\n### Read the Voice Blueprint First\n\nIf `marketing-context.md` is available: read the brand voice section and writing examples. If not, ask for one example of content this brand loves. One. Then extract the patterns from it.\n\n**What to extract from a voice example:**\n- Sentence length preference (short punchy vs. longer flowing?)\n- Formality level (contractions? slang? industry jargon?)\n- Use of humor (dry wit? self-deprecating? none?)\n- Relationship stance (peer-to-peer? expert-to-student? provocateur?)\n- Signature phrases or patterns\n\nSee [references/voice-techniques.md](references/voice-techniques.md) for specific techniques for each voice type.\n\n### Voice Injection Techniques\n\n**1. Personal Anecdotes**\nEven branded content gets more credible when grounded in experience. \"We saw this firsthand when building X\" is worth more than any study citation.\n\n**2. Direct Address**\nTalk to the reader as \"you.\" Not \"users\" or \"teams\" or \"organizations.\" You.\n\n**3. Opinions Without Apology**\nState your position. \"We think the industry is wrong about this\" is more credible than \"there are various perspectives.\" Take the side.\n\n**4. The Aside**\nA brief parenthetical that shows the brand knows more than it's saying. \"This also affects API performance, but that's a separate rabbit hole.\"\n\n**5. Rhythm Signature**\nEvery brand has a rhythm. Some write in short staccato bursts. Some write long, winding sentences that spiral back on themselves. Find the rhythm from the examples and apply it consistently.\n\n### Before / After Example\n\n**Before (AI-generated):**\n> It is crucial to leverage your existing customer data in order to effectively navigate the competitive landscape. Furthermore, by implementing a robust onboarding strategy, organizations can ensure that users achieve maximum value from the product and reduce churn significantly.\n\n**After (humanized):**\n> Here's the thing nobody says out loud: most SaaS companies have the data to fix their churn problem. They just don't look at it until after customers leave.\n>\n> Your activation funnel is in there. Your best cohorts, your worst, the moment the drop-off happens. You don't need another tool — you need someone to stop ignoring what the tool is already showing you.\n>\n> Nail onboarding first. Everything else is downstream.\n\nWhat changed:\n- Removed: \"crucial,\" \"leverage,\" \"navigate,\" \"robust,\" \"ensure,\" \"significantly,\" \"furthermore\"\n- Added: direct address, specific accusation (\"what the tool is already showing you\"), short-sentence punch at the end\n- Changed: passive recommendations → active point of view\n\n---\n\n## Proactive Triggers\n\nFlag these without being asked:\n\n- **AI fingerprint density too high** — If the piece has 10+ AI tells per 500 words, a patch job won't work. Flag that the piece needs a full rewrite, not an edit. Trying to polish a piece that's 80% AI patterns produces AI patterns with nicer words.\n- **Voice context missing** — If `marketing-context.md` doesn't exist and the user hasn't given voice guidance, pause before injecting voice. Ask for one example. Guessing the voice and being wrong wastes everyone's time.\n- **Specificity gap** — If the piece makes 5+ vague claims with zero data or attribution, flag it to the user. You can make the prose flow better, but you can't invent specific proof. They need to provide it.\n- **Tone mismatch after humanizing** — If the piece is now genuinely human but sounds like a different brand than everything else the client publishes, flag it. Consistency matters as much as quality.\n- **Over-editing risk** — If the original content has one or two genuinely good paragraphs buried in the AI mush, flag them before rewriting. Don't accidentally destroy the good parts.\n\n---\n\n## Output Artifacts\n\n| When you ask for... | You get... |\n|---|---|\n| AI audit | Annotated version of the draft with each AI pattern flagged, severity score, and count by category |\n| Humanized draft | Full rewrite with AI patterns removed, rhythm varied, specificity improved |\n| Voice injection | Annotated draft with brand voice applied — specific changes called out so you can learn the pattern |\n| Before/after comparison | Side-by-side view of key paragraphs showing what changed and why |\n| Humanity score | Run `scripts/humanizer_scorer.py` — 0-100 score with breakdown by signal type |\n\n---\n\n## Communication\n\nAll output follows the structured standard:\n- **Bottom line first** — answer before explanation\n- **What + Why + How** — every finding includes all three\n- **Actions have owners and deadlines** — no \"you might want to consider\"\n- **Confidence tagging** — 🟢 verified pattern / 🟡 medium / 🔴 assumed based on limited voice context\n\nWhen auditing: name the pattern → explain why it reads as AI → give the specific fix. Not \"this sounds robotic.\" Say: \"Paragraph 4 opens with 'It is important to note that' — this is a pure hedge. Cut it. Start with the actual note.\"\n\n---\n\n## Related Skills\n\n- **content-production**: Use to produce the initial draft. Run content-humanizer after drafting, before the SEO optimization pass.\n- **copywriting**: Use for conversion copy — landing pages, CTAs, headlines. content-humanizer works on longer-form pieces; copywriting handles short punchy copy with different principles.\n- **content-strategy**: Use when deciding what content to create. NOT for voice or draft execution.\n- **ai-seo**: Use after humanizing, to optimize for AI search citation. Human-sounding content gets cited more — but it still needs structure to get extracted.\n\nFile v1.0.0:_meta.json\n\n{\n  \"ownerId\": \"kn73z7zhz92tv6skpnmpy5pdfh86j68x\",\n  \"slug\": \"content-humanizer\",\n  \"version\": \"1.0.0\",\n  \"publishedAt\": 1781827386570\n}\n\nFile v1.0.0:references/ai-tells-checklist.md\n\n# AI Tells Checklist\n\nA comprehensive reference for detecting AI-generated or AI-assisted writing patterns. Use this during Mode 1 (Detect) to audit content before editing.\n\nRate each finding: 🔴 Critical (rewrites required) / 🟡 Medium (edits needed) / 🟢 Minor (polish)\n\n---\n\n## Category 1: Overused Vocabulary\n\nThese words appear in AI output at 5-20x the frequency of human writing. One instance is fine. Multiple instances in a single piece is a tell.\n\n### Red-flag verbs (overused)\n| Word | Problem | What humans say instead |\n|---|---|---|\n| delve / delve into | Pretentious filler | look at, dig into, explore, examine |\n| leverage | Jargon for \"use\" | use, apply, put to work, capitalize on |\n| foster | Formal filler | build, develop, encourage, create |\n| facilitate | Bureaucratic filler | help, make easier, allow, support |\n| navigate | Overused metaphor | handle, manage, work through, deal with |\n| ensure | Empty guarantee | check that, make sure, verify |\n| utilize | Formal for \"use\" | just say \"use\" |\n| prioritize | Often redundant | just say what to do first |\n| streamline | Vague improvement promise | be specific about what gets faster/easier |\n\n### Red-flag adjectives (overused)\n| Word | Problem | What humans say instead |\n|---|---|---|\n| crucial / vital / pivotal | Overloaded intensifiers | just show why it matters |\n| robust | Vague positive | specific: \"handles X load,\" \"covers Y cases\" |\n| comprehensive | Overpromise | specific: \"covers the 5 most common...\" |\n| innovative | Meaningless | say what it actually does differently |\n| holistic | Buzzword | say what it covers |\n| seamless | Overused product adjective | describe the actual experience |\n| cutting-edge | Dated marketing | say what's new or different specifically |\n| dynamic | Filler adjective | usually just delete it |\n\n### Red-flag nouns (overused)\n| Word | Problem | What humans say instead |\n|---|---|---|\n| landscape | Overused metaphor | \"how X works today,\" \"the state of X\" |\n| ecosystem | Overused for \"industry\" or \"community\" | be specific |\n| framework | Often vague | name the specific framework or approach |\n| paradigm | Academic filler | usually replaceable with \"approach\" or \"way of thinking\" |\n| synergy | Corporate cliché | delete or replace with what actually happens |\n\n---\n\n## Category 2: Hedging and Qualification Patterns\n\nHumans hedge. AI hedges compulsively. The difference is frequency and context.\n\n### Opening hedges (usually safe to cut)\n- \"It's important to note that...\"\n- \"It's worth mentioning that...\"\n- \"It should be noted that...\"\n- \"Needless to say...\"\n- \"It goes without saying...\"\n- \"Of course, ...\"\n- \"Naturally, ...\"\n\n### Mid-sentence hedges (examine each)\n- \"In many cases\" / \"In most instances\" / \"In certain scenarios\"\n- \"Generally speaking,\" / \"For the most part,\"\n- \"This may vary depending on...\"\n- \"Results may differ based on...\"\n- \"While this isn't always the case...\"\n\n**Diagnostic:** If the hedge is protecting a claim that should just be a claim, cut the hedge and state the claim. If the hedge reflects genuine uncertainty, keep it — but make the uncertainty specific: \"I don't have data for this, but based on [context]...\"\n\n### Vague authority claims (replace with specifics)\n- \"Studies show...\" → \"A 2023 McKinsey study of 400 SaaS companies showed...\"\n- \"Research suggests...\" → same — name the research\n- \"Many companies...\" → \"HubSpot, Slack, and several bootstrapped SaaS founders we've talked to...\"\n- \"Experts agree...\" → name one expert\n- \"It has been shown that...\" → by whom, when, where\n\n---\n\n## Category 3: Structural Patterns\n\n### The SEEB paragraph (most common)\nPattern: Statement → Explanation → Example → Bridge to next topic\nWhen every paragraph follows this exact structure, the writing reads like a machine assembled it. Because it was.\n\n**Fix:** Mix in fragments. Questions. Short paragraphs. Asides. Let sections breathe differently.\n\n### Parallelism overload\nAI loves parallel structure. Three-item lists. Four-item lists. Everything in threes. Alliteration sometimes.\n\nOccasional parallelism is powerful. Three consecutive bulleted lists of three items each is a tell.\n\n**Fix:** Break the rhythm. Some sections need bullets. Most don't.\n\n### The summary conclusion\nAI conclusions restate the introduction. Paragraph by paragraph. \"In this guide, we covered X, Y, and Z. By applying these strategies, you can achieve [thing from the intro].\"\n\nHuman conclusions either add something (a new angle, an honest admission, a call to action that feels earned) or they nail the exit line and stop.\n\n### Symmetric section lengths\nEvery H2 section is roughly the same length. ~300 words. Every one. That's a machine maintaining consistency. Humans have opinions — some things deserve more space, some less.\n\n---\n\n## Category 4: Punctuation and Formatting Tells\n\n### Em-dash frequency 🟡\nOne or two em-dashes per piece: fine. Three per page: suspicion. Five+: AI fingerprint.\n\nThe em-dash is AI's favorite way to add subordinate clauses — like this — because it sounds sophisticated — and it learned this from a lot of well-written text — so now it can't stop.\n\n**Fix:** Replace most with periods. Break into shorter sentences.\n\n### Colon-then-list patterns 🟢\nAI frequently: introduces a list: with: colons. Multiple times per section. Lists are useful. But if every section has a colon-introducing list, it's mechanical.\n\n### Excessive bold 🟢\nAI sometimes bolds every important-sounding phrase in a paragraph. **This results** in **too many** phrases being **highlighted for emphasis** when the emphasis **dilutes itself**.\n\nBold should be used sparingly — for the one thing that matters most in a section, not for three to five things per paragraph.\n\n---\n\n## Category 5: Tonal Tells\n\n### False warmth 🟡\n\"We hope this guide has been helpful in your journey to...\" \n\"We trust that you've found valuable insights in...\"\n\"It's our sincere hope that these strategies will empower you to...\"\n\nNo one talks like this. It's the corporate newsletter voice. Cut it.\n\n### Emotional escalation without basis 🟡\nAI sometimes starts clinical and then, near the end, gets unexpectedly warm and inspirational. \"Now that you have these tools, you can transform your business and achieve your goals.\" The warmth wasn't earned by the preceding content.\n\n### Artificial enthusiasm 🟢\n\"Exciting developments,\" \"fascinating case study,\" \"incredible opportunity\" — used to simulate human engagement. Usually reads as hollow because it's disconnected from actual content that earns those descriptors.\n\n---\n\n## Quick Audit Scoring\n\nFor a 1,000-word piece, count:\n\n| Signal | Count | Severity Threshold |\n|---|---|---|\n| Red-flag vocabulary words | | >5 = 🔴 |\n| Opening hedges | | >2 = 🔴 |\n| Vague authority claims (\"studies show\") | | >2 = 🔴 |\n| Em-dashes | | >4 = 🟡 |\n| SEEB paragraphs (all follow same structure) | | >60% of paragraphs = 🔴 |\n| Instances of false warmth | | >1 = 🟡 |\n\n**Scoring:**\n- 0-5 total flags: Light edit (Mode 2 quick pass)\n- 6-12 total flags: Full humanize pass (Mode 2 complete)\n- 12+ total flags: Full rewrite recommended (Mode 1 audit → complete Mode 2 → Mode 3)\n\nFile v1.0.0:references/voice-techniques.md\n\n# Voice Techniques Reference\n\nTechniques for injecting authentic brand voice into content. This is the Mode 3 playbook — after you've removed AI patterns (Mode 2), these techniques put the brand's specific personality in.\n\n---\n\n## Step 1: Extract the Voice Profile\n\nBefore writing anything, extract the voice from examples. Ask for one piece of content the brand considers representative. Then answer these questions about it:\n\n| Question | What to look for |\n|---|---|\n| Average sentence length? | Count words per sentence across 10 sentences. ≤15 = punchy. ≥25 = flowing. |\n| Formality level? | Count contractions (\"it's\" vs \"it is\"). Count first-person. Count slang. |\n| Use of humor? | Dry wit (unexpected juxtapositions). Self-deprecating (acknowledges own limitations). Provocateur (picks fights). None. |\n| Relationship to reader? | Peer (we're both figuring this out). Expert (I know, you're learning). Challenger (you're probably wrong about this). |\n| Signature phrases? | Phrases or constructions that appear more than once. That's the brand's verbal tic. |\n| What do they avoid? | Listen for what's conspicuously absent. |\n\nDocument this as a voice profile before editing. Reference it throughout.\n\n---\n\n## The Core Voice Types\n\n### Voice Type 1: The Direct Expert\n\n**Profile:** Short sentences. Confident claims. No hedging. No qualifications unless the qualification is the point. Reads like advice from the smartest person in the room who also has the least patience.\n\n**Natural habitat:** Developer tools, technical documentation, no-fluff B2B brands, opinionated founders.\n\n**Techniques:**\n- Lead every section with the conclusion, not the setup\n- Make claims without softening them: \"Most onboarding is broken\" not \"Many onboarding flows may have room for improvement\"\n- Use \"you\" relentlessly — direct address, always\n- Short paragraphs. Often 1-2 sentences.\n- When listing, use the smallest number of words possible\n- No rhetorical questions unless you're about to immediately answer them\n- Swear if the brand does (a well-placed \"this is genuinely terrible\" signals authenticity)\n\n**Before:** \"It's important to consider that implementing an effective onboarding strategy can have significant positive impacts on user retention rates.\"\n\n**After:** \"Fix your onboarding. It's where you're losing them.\"\n\n---\n\n### Voice Type 2: The Thoughtful Peer\n\n**Profile:** Longer sentences when the thought requires it. Occasional vulnerability. Shares the thinking, not just the conclusion. Treats the reader as someone at roughly the same level.\n\n**Natural habitat:** B2B SaaS with community focus, content brands, newsletter-first companies, founder blogs.\n\n**Techniques:**\n- Show the reasoning: \"Here's how I think about this...\" not just the conclusion\n- Acknowledge what you don't know: \"I don't have solid data on this, but my read is...\"\n- Share the mistake before the lesson: \"We tried X first and it failed. Here's what we learned.\"\n- Longer transitions: connect ideas, don't just jump to the next point\n- Questions that earn their keep — rhetorical only when you're genuinely pointing at a tension\n- First person plural (\"we\") where appropriate — but only if there actually is a \"we\"\n\n**Before:** \"Companies should invest in customer success to improve retention.\"\n\n**After:** \"I spent two years thinking retention was a marketing problem. It wasn't. Every company we talked to that had great retention had one thing in common: they treated customer success like a product team, not a support team. The product mindset made all the difference.\"\n\n---\n\n### Voice Type 3: The Provocateur\n\n**Profile:** Challenges received wisdom. Takes positions most people avoid. Enjoys being right when everyone else was wrong. Not confrontational for its own sake — has actual conviction behind the provocation.\n\n**Natural habitat:** Opinionated SaaS, contrarian analyst voices, certain agency brands, fast-growing startups trying to differentiate.\n\n**Techniques:**\n- Open with the unpopular claim: \"Most SEO advice is wrong\" not \"There are many perspectives on SEO\"\n- Name the thing people don't say: \"The reason nobody talks about X is because...\"\n- Acknowledge the argument against your position before making yours stronger\n- Use \"actually\" strategically: \"What actually happens is...\"\n- When everyone agrees on something, be specific about what they're getting wrong\n- Let the evidence land without fanfare — state the uncomfortable number, then wait\n\n**Before:** \"There are various approaches to content marketing that teams may find valuable.\"\n\n**After:** \"Your competitors' content marketing budgets are wasted. The average B2B blog post gets 300 views and zero conversions. The top 10% of pieces generate 90% of the value. Most teams don't know which 10% they have — which means they're spending 90% of their budget on content that will never pay back.\"\n\n---\n\n### Voice Type 4: The Enthusiastic Practitioner\n\n**Profile:** Genuine excitement about the craft. Energized by the detail. The person who could talk about their domain for hours and you'd actually enjoy it.\n\n**Natural habitat:** Creative tools, marketing tech, community platforms, brands with a strong practitioner user base.\n\n**Techniques:**\n- Allow enthusiasm to come through without forcing it: \"This is the part I love about cohort analysis...\"\n- Share the unexpected detail that only a practitioner would notice\n- Use specific terminology confidently — don't over-explain to practitioners\n- \"The thing is...\" as a transition — signals you're about to share the real insight\n- Parenthetical asides that reveal depth: \"(And yes, I know about [edge case] — that's a separate problem)\"\n- Let the complexity show when the complexity is interesting\n\n**Before:** \"Segmenting your audience allows for more targeted communication.\"\n\n**After:** \"Here's where audience segmentation gets genuinely interesting. Most people segment by company size or industry. That's fine. But the cohort that always outperforms is behavioral — people who hit your activation event within 72 hours. That cohort has 2-3x the LTV of the same-company-size cohort that took two weeks. The behavior tells you more than the firmographic ever will.\"\n\n---\n\n## Specific Techniques Regardless of Voice Type\n\n### The Personal Anecdote Injection\n\nEvery brand can use a personal anecdote — even B2B technical brands. The format:\n\n**Setup:** \"When we [did X thing in context]...\"\n**What happened:** \"[Specific unexpected thing occurred].\"\n**The lesson:** \"[The principle that explains it].\"\n\nThe anecdote doesn't have to be dramatic. It just has to be real and specific. \"When we rebuilt our onboarding flow in 2023, we expected the copy to be the problem. It was the loading time.\"\n\n### The Honest Admission\n\nInstantly humanizing. Rarely done by brands because it feels like weakness. But readers trust brands that admit:\n- What they got wrong\n- What they don't know\n- What they tried that failed\n- What surprised them\n\nThe format: \"We thought X. We were wrong. What actually happened was Y.\"\n\n### The Contrarian Setup\n\nWorks for any voice type. Setup: state the conventional wisdom. Then undercut it.\n\n\"Everyone says [thing]. The data shows [opposite thing].\"\n\nOr more subtle: \"The conventional wisdom is [X]. That's mostly true. Except when [specific condition] — and that's where it gets interesting.\"\n\n### Sentence Rhythm Signature\n\nAfter analyzing the brand voice, identify the rhythm pattern and apply it:\n\n- **Staccato brand:** Short sentence. Shorter. Done.\n- **Flowing brand:** Content that allows ideas to build on each other, connect, and arrive somewhere richer than where they started.\n- **Mixed cadence (most natural):** Long sentence establishing context. Short punch. Then another long one that goes somewhere. Then: done.\n\nOnce you know the pattern, apply it consistently throughout the piece.\n\n### The Earned Ending\n\nThe last sentence of a piece should land. Not summarize. Not encourage. Land.\n\nTechniques:\n- **The hard cut:** State the core truth one final time, nakedly. No wrap-up.\n- **The reversal:** Setup expectation, then subvert it in the last line.\n- **The action:** Name the single most important thing to do next. One thing.\n- **The honest admission:** \"We're still figuring this out too.\"\n\nAvoid: \"By following these steps, you'll be well on your way to achieving your goals.\" Nobody reads that. Nobody trusts it.\n\n---\n\n## Voice Injection Quality Check\n\nAfter applying voice:\n- [ ] Read the piece aloud — does it sound like a person you'd actually follow?\n- [ ] Is every paragraph distinguishable from the others? (Not just same rhythm, same length, same pattern)\n- [ ] Is there at least one moment where the brand's specific personality is unmistakable?\n- [ ] Would someone who knows this brand's other content recognize this as theirs?\n- [ ] Is there at least one place where the brand says something most brands wouldn't?\n\nFile v1.0.0:skill-card.md\n\n## Description: <br>\nContent Humanizer helps agents audit AI-sounding marketing copy, rewrite it with more natural rhythm and specificity, and adapt it to a brand voice. <br>\n\nThis skill is ready for commercial/non-commercial use. <br>\n\n## Publisher: <br>\n[paudyyin](https://clawhub.ai/user/paudyyin) <br>\n\n### License/Terms of Use: <br>\nMIT-0 <br>\n\n\n## Use Case: <br>\nMarketing and content teams use this skill to turn robotic or generic drafts into more credible human-facing copy. It supports AI-pattern audits, full rewrites, voice injection, before-and-after comparisons, and heuristic humanity scoring. <br>\n\n### Deployment Geography for Use: <br>\nGlobal <br>\n\n## Known Risks and Mitigations: <br>\nRisk: The humanity score is a heuristic and should not be treated as definitive authorship detection. <br>\nMitigation: Use the score as an editing signal only, and rely on human review before making authorship, quality, or publication decisions. <br>\nRisk: Brand voice context or pasted drafts may contain sensitive marketing details. <br>\nMitigation: Provide only the context needed for the edit and review rewritten or annotated output before sharing it outside the intended audience. <br>\nRisk: Humanizing a draft can change factual claims, tone, or useful original passages. <br>\nMitigation: Review the final copy against source material, preserve strong original sections when flagged, and provide evidence for specific claims rather than inventing proof. <br>\n\n\n## Reference(s): <br>\n- [Content Humanizer on ClawHub](https://clawhub.ai/paudyyin/content-humanizer) <br>\n- [AI Tells Checklist](references/ai-tells-checklist.md) <br>\n- [Voice Techniques Reference](references/voice-techniques.md) <br>\n\n\n## Skill Output: <br>\n**Output Type(s):** [text, markdown, shell commands, guidance] <br>\n**Output Format:** [Markdown with rewritten copy, annotated findings, comparisons, and optional local scoring output] <br>\n**Output Parameters:** [1D] <br>\n**Other Properties Related to Output:** [May ask for brand voice context or read marketing-context.md when present; the local scorer reports a 0-100 heuristic humanity score with signal breakdowns.] <br>\n\n## Skill Version(s): <br>\n1.0.0 (source: server release metadata, package.json, and SKILL.md metadata) <br>\n\n## Ethical Considerations: <br>\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment. <br>\n\nFile v1.0.0:package.json\n\n{\r\n    \"name\":  \"content-humanizer\",\r\n    \"version\":  \"1.0.0\",\r\n    \"description\":  \"Makes AI-generated content sound genuinely human — not just cleaned up, but alive. Use when content feels robotic, uses too many AI clichés, lacks personality, or reads like it was written by committee. Triggers: \\u0027this sounds like AI\\u0027, \\u0027make it more human\\u0027, \\u0027add personality\\u0027, \\u0027it feels generic\\u0027, \\u0027sounds robotic\\u0027, \\u0027fix AI writing\\u0027, \\u0027inject our voice\\u0027. NOT for initial content creation (use content-production). NOT for SEO optimization (use content-production Mode 3).\"\r\n}","readmeExcerpt":"Skill: Content Humanizer Owner: paudyyin Summary: Transform AI-generated content into genuinely human-sounding text. Use when content feels robotic, uses AI cliches, lacks personality, or reads generically.... Tags: latest:1.0.1 Version history: v1.0.1 | 2026-07-17T00:28:43.032Z | user Optimize description: English verb-first + Use when format, No-Op Test cleanup (-14% lines) v1.0.0 | 2026-06-19T00:03:06.570Z | auto ","codeSnippets":[],"executableExamples":[],"parameters":null,"dependencies":[],"permissions":[],"extractedFiles":[{"path":"SKILL.md","content":"---\nname: \"content-humanizer\"\ndescription: \"Transform AI-generated content into genuinely human-sounding text. Use when content feels robotic, uses AI cliches, lacks personality, or reads generically. NOT for initial content creation or SEO optimization.\"\nlicense: MIT\nmetadata:\n  version: 1.0.0\n  author: Alireza Rezvani\n  category: marketing\n  updated: 2026-03-06\n---\n\n# Content Humanizer\n\nYou are an expert in authentic writing and brand voice. Your goal is to transform content that reads like it was generated by a machine �?even when it technically was �?into writing that sounds like a real person with real opinions, real experience, and real stakes in what they're saying.\n\nThis is not a cleaning service. You're not just removing \"delve\" and calling it a day. You're rebuilding the voice from the ground up.\n\n## Before Starting\n\n**Check for context first:**\nIf `marketing-context.md` exists, read it. It contains brand voice guidelines, writing examples, and the specific tone this brand uses. That context is your voice blueprint. Use it �?don't improvise a voice when the brief already defines one.\n\nGather what you need before starting:\n\n### What you need\n- **The content** �?paste the draft to humanize\n- **Brand voice notes** �?if no `marketing-context.md`, ask: \"Is your voice direct/casual/technical/irreverent? Give me one example of writing you love.\"\n- **Audience** �?who reads this? (This changes what \"human\" sounds like)\n- **Goal** �?what should this piece do? (Knowing the goal tells you how much personality is appropriate)\n\nOne question if needed: \"Before I rewrite this, give me an example of content you've written or read that felt right. Specific is better than descriptive.\"\n\n## How This Skill Works\n\nThree modes. Run them in sequence for a full transformation, or jump to the one you need:\n\n### Mode 1: Detect �?AI Pattern Analysis\nAudit the content for AI tells. Name what's wrong and why before fixing anything. This is diagnostic �?not editorial.\n\n### Mode 2: Humanize �?Pattern Removal and Rhythm Fix\nStrip the AI patterns. Fix sentence rhythm. Replace generic with specific. The content starts sounding like a person.\n\n### Mode 3: Voice Injection �?Brand Character\nNow that the generic is gone, inject the brand's specific personality. This is where \"human\" becomes *your brand's* human.\n\nRun all three in one pass when you have enough context. Split them when the client needs to see the audit before you edit.\n\n---\n\n## Mode 1: Detect �?AI Pattern Analysis\n\nScan the content for these categories. Score severity: 🔴 critical (kills credibility) / 🟡 medium (softens impact) / 🟢 minor (polish only).\n\nSee [references/ai-tells-checklist.md](references/ai-tells-checklist.md) for the comprehensive detection list.\n\n### The Core AI Tell Categories\n\n**1. Overused Filler Words** 🔴\nThe model loves certain words because they appear frequently in its training data. Flag these on sight:\n- \"delve,\" \"delve into,\" \"delve deeper\"\n- \"landscape\" (as in \"the current AI lan"},{"path":"_meta.json","content":"{\n  \"ownerId\": \"kn73z7zhz92tv6skpnmpy5pdfh86j68x\",\n  \"slug\": \"content-humanizer\",\n  \"version\": \"1.0.1\",\n  \"publishedAt\": 1784248123032\n}"},{"path":"references/ai-tells-checklist.md","content":"# AI Tells Checklist\n\nA comprehensive reference for detecting AI-generated or AI-assisted writing patterns. Use this during Mode 1 (Detect) to audit content before editing.\n\nRate each finding: 🔴 Critical (rewrites required) / 🟡 Medium (edits needed) / 🟢 Minor (polish)\n\n---\n\n## Category 1: Overused Vocabulary\n\nThese words appear in AI output at 5-20x the frequency of human writing. One instance is fine. Multiple instances in a single piece is a tell.\n\n### Red-flag verbs (overused)\n| Word | Problem | What humans say instead |\n|---|---|---|\n| delve / delve into | Pretentious filler | look at, dig into, explore, examine |\n| leverage | Jargon for \"use\" | use, apply, put to work, capitalize on |\n| foster | Formal filler | build, develop, encourage, create |\n| facilitate | Bureaucratic filler | help, make easier, allow, support |\n| navigate | Overused metaphor | handle, manage, work through, deal with |\n| ensure | Empty guarantee | check that, make sure, verify |\n| utilize | Formal for \"use\" | just say \"use\" |\n| prioritize | Often redundant | just say what to do first |\n| streamline | Vague improvement promise | be specific about what gets faster/easier |\n\n### Red-flag adjectives (overused)\n| Word | Problem | What humans say instead |\n|---|---|---|\n| crucial / vital / pivotal | Overloaded intensifiers | just show why it matters |\n| robust | Vague positive | specific: \"handles X load,\" \"covers Y cases\" |\n| comprehensive | Overpromise | specific: \"covers the 5 most common...\" |\n| innovative | Meaningless | say what it actually does differently |\n| holistic | Buzzword | say what it covers |\n| seamless | Overused product adjective | describe the actual experience |\n| cutting-edge | Dated marketing | say what's new or different specifically |\n| dynamic | Filler adjective | usually just delete it |\n\n### Red-flag nouns (overused)\n| Word | Problem | What humans say instead |\n|---|---|---|\n| landscape | Overused metaphor | \"how X works today,\" \"the state of X\" |\n| ecosystem | Overused for \"industry\" or \"community\" | be specific |\n| framework | Often vague | name the specific framework or approach |\n| paradigm | Academic filler | usually replaceable with \"approach\" or \"way of thinking\" |\n| synergy | Corporate cliché | delete or replace with what actually happens |\n\n---\n\n## Category 2: Hedging and Qualification Patterns\n\nHumans hedge. AI hedges compulsively. The difference is frequency and context.\n\n### Opening hedges (usually safe to cut)\n- \"It's important to note that...\"\n- \"It's worth mentioning that...\"\n- \"It should be noted that...\"\n- \"Needless to say...\"\n- \"It goes without saying...\"\n- \"Of course, ...\"\n- \"Naturally, ...\"\n\n### Mid-sentence hedges (examine each)\n- \"In many cases\" / \"In most instances\" / \"In certain scenarios\"\n- \"Generally speaking,\" / \"For the most part,\"\n- \"This may vary depending on...\"\n- \"Results may differ based on...\"\n- \"While this isn't always the case...\"\n\n**Diagnostic:** If the hedge is protecting a claim that should just be a claim"},{"path":"references/voice-techniques.md","content":"# Voice Techniques Reference\n\nTechniques for injecting authentic brand voice into content. This is the Mode 3 playbook — after you've removed AI patterns (Mode 2), these techniques put the brand's specific personality in.\n\n---\n\n## Step 1: Extract the Voice Profile\n\nBefore writing anything, extract the voice from examples. Ask for one piece of content the brand considers representative. Then answer these questions about it:\n\n| Question | What to look for |\n|---|---|\n| Average sentence length? | Count words per sentence across 10 sentences. ≤15 = punchy. ≥25 = flowing. |\n| Formality level? | Count contractions (\"it's\" vs \"it is\"). Count first-person. Count slang. |\n| Use of humor? | Dry wit (unexpected juxtapositions). Self-deprecating (acknowledges own limitations). Provocateur (picks fights). None. |\n| Relationship to reader? | Peer (we're both figuring this out). Expert (I know, you're learning). Challenger (you're probably wrong about this). |\n| Signature phrases? | Phrases or constructions that appear more than once. That's the brand's verbal tic. |\n| What do they avoid? | Listen for what's conspicuously absent. |\n\nDocument this as a voice profile before editing. Reference it throughout.\n\n---\n\n## The Core Voice Types\n\n### Voice Type 1: The Direct Expert\n\n**Profile:** Short sentences. Confident claims. No hedging. No qualifications unless the qualification is the point. Reads like advice from the smartest person in the room who also has the least patience.\n\n**Natural habitat:** Developer tools, technical documentation, no-fluff B2B brands, opinionated founders.\n\n**Techniques:**\n- Lead every section with the conclusion, not the setup\n- Make claims without softening them: \"Most onboarding is broken\" not \"Many onboarding flows may have room for improvement\"\n- Use \"you\" relentlessly — direct address, always\n- Short paragraphs. Often 1-2 sentences.\n- When listing, use the smallest number of words possible\n- No rhetorical questions unless you're about to immediately answer them\n- Swear if the brand does (a well-placed \"this is genuinely terrible\" signals authenticity)\n\n**Before:** \"It's important to consider that implementing an effective onboarding strategy can have significant positive impacts on user retention rates.\"\n\n**After:** \"Fix your onboarding. It's where you're losing them.\"\n\n---\n\n### Voice Type 2: The Thoughtful Peer\n\n**Profile:** Longer sentences when the thought requires it. Occasional vulnerability. Shares the thinking, not just the conclusion. Treats the reader as someone at roughly the same level.\n\n**Natural habitat:** B2B SaaS with community focus, content brands, newsletter-first companies, founder blogs.\n\n**Techniques:**\n- Show the reasoning: \"Here's how I think about this...\" not just the conclusion\n- Acknowledge what you don't know: \"I don't have solid data on this, but my read is...\"\n- Share the mistake before the lesson: \"We tried X first and it failed. Here's what we learned.\"\n- Longer transitions: connect ideas, don't just j"},{"path":"skill-card.md","content":"## Description:\n\nTransforms AI-generated content into genuinely human-sounding text when drafts feel robotic, use AI cliches, lack personality, or read generically, and is not intended for initial content creation or SEO optimization.\n\nThis skill is ready for commercial/non-commercial use.\n\n## Publisher:\n\n[paudyyin](https://clawhub.ai/user/paudyyin)\n\n### License/Terms of Use:\n\nMIT-0\n\n## Use Case:\n\nMarketing and content teams use this skill to audit AI-sounding drafts, rewrite them with more natural rhythm and specificity, and apply a supplied brand voice. It is intended for improving existing content rather than creating first drafts or SEO optimization.\n\n### Deployment Geography for Use:\n\nGlobal\n\n## Known Risks and Mitigations:\n\nRisk: The skill may transform supplied drafts and optional brand-voice context, which can include sensitive unpublished marketing content.\n\nMitigation: Use appropriate input handling for confidential drafts and review the rewritten output before sharing or publishing.\n\nRisk: Humanizing content can change tone, emphasis, or factual specificity in ways that misrepresent the intended message.\n\nMitigation: Compare the output against the original brief, supplied evidence, and brand voice before publication.\n\nRisk: The skill documentation includes a fallback to Chinese when language detection fails, which may produce the wrong language for some drafts.\n\nMitigation: Specify the target language when the source content is short, mixed-language, or ambiguous.\n\n## Reference(s):\n\n- [ClawHub Skill Page](https://clawhub.ai/paudyyin/skills/content-humanizer)\n- [AI Tells Checklist](references/ai-tells-checklist.md)\n- [Voice Techniques Reference](references/voice-techniques.md)\n\n## Skill Output:\n\n**Output Type(s):** [text, markdown, shell commands, guidance]\n\n**Output Format:** [Markdown with optional plaintext reports and JSON scorer output]\n\n**Output Parameters:** [1D]\n\n**Other Properties Related to Output:** [May produce an AI-pattern audit, humanized rewrite, voice-injected draft, before/after comparison, or local humanity-score report.]\n\n## Skill Version(s):\n\n1.0.1 (source: server release metadata; artifact frontmatter and package.json report 1.0.0)\n\n## Ethical Considerations:\n\nUsers should evaluate whether this skill is appropriate for their environment, review any generated or modified files before relying on them, and apply their organization's safety, security, and compliance requirements before deployment."}],"languages":[],"docsSourceLabel":"CLAWHUB","editorialOverview":"Transform AI-generated content into genuinely human-sounding text. Use when content feels robotic, uses AI cliches, lacks personality, or reads generically.... Skill: Content Humanizer Owner: paudyyin Summary: Transform AI-generated content into genuinely human-sounding text. Use when content feels robotic, uses AI cliches, lacks personality, or reads generically.... Tags: latest:1.0.1 Version history: v1.0.1 | 2026-07-17T00:28:43.032Z | user Optimize description: English verb-first + Use when format, No-Op Test cleanup (-14% lines) v1.0.0 | 2026-06-19T00:03:06.570Z | auto","editorialQuality":{"score":100,"threshold":65,"status":"ready","wordCount":2025,"uniquenessScore":48,"reasons":[]}},"media":{"evidence":{"source":"no-media","verified":false,"confidence":"low","updatedAt":"2026-10-11T21:01:17.577Z","emptyReason":"No screenshots, media assets, or demo links are available."},"primaryImageUrl":null,"mediaAssetCount":0,"assets":[],"demoUrl":null},"ownerResources":{"evidence":{"source":"unclaimed","verified":false,"confidence":"low","updatedAt":"2026-10-11T21:01:17.577Z","emptyReason":"This page has not been claimed by the agent owner."},"hasCustomPage":false,"customPageUpdatedAt":null,"customLinks":[],"structuredLinks":{"docsUrl":null,"demoUrl":null,"supportUrl":null,"pricingUrl":null,"statusUrl":null},"customPage":null},"relatedAgents":{"evidence":{"source":"protocol-neighbors","verified":false,"confidence":"medium","updatedAt":"2026-10-13T21:57:20.441Z","emptyReason":null},"items":[{"id":"8ebccd8e-3863-4187-8355-c3f14e1f9edf","entityType":"agent","canonicalPath":"/agent/iofficeai-aionui","slug":"iofficeai-aionui","name":"AionUi","description":"Free, local, open-source 24/7 Cowork app and OpenClaw for Gemini CLI, Claude Code, Codex, OpenCode, Qwen Code, Goose CLI, Auggie, and more | 🌟 Star if you like it!","url":"https://github.com/iOfficeAI/AionUi","homepage":"https://www.aionui.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-10-09T19:11:12.944Z","createdAt":"2026-02-25T03:38:16.584Z","downloads":null},{"id":"b917f68a-ebff-438e-84f8-3f4b2494c0bc","entityType":"agent","canonicalPath":"/agent/activepieces-activepieces","slug":"activepieces-activepieces","name":"activepieces","description":"AI Agents & MCPs & AI Workflow Automation • (~400 MCP servers for AI agents) • AI Automation / AI Agent with MCPs • AI Workflows & AI Agents • MCPs for AI Agents","url":"https://github.com/activepieces/activepieces","homepage":"https://www.activepieces.com","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-15T02:22:12.426Z","createdAt":"2026-02-25T03:38:12.412Z","downloads":null},{"id":"5cb26759-3a39-483f-94cf-276a98c13bb8","entityType":"agent","canonicalPath":"/agent/cherryhq-cherry-studio","slug":"cherryhq-cherry-studio","name":"cherry-studio","description":"AI productivity studio with smart chat, autonomous agents, and 300+ assistants. Unified access to frontier LLMs","url":"https://github.com/CherryHQ/cherry-studio","homepage":"https://cherry-ai.com","source":"GITHUB_REPOS","protocols":["MCP","OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-04-11T14:38:40.986Z","createdAt":"2026-02-25T03:38:19.379Z","downloads":null},{"id":"6f6582d0-5d76-4f0f-b81d-86520247950b","entityType":"agent","canonicalPath":"/agent/copilotkit-copilotkit","slug":"copilotkit-copilotkit","name":"CopilotKit","description":"The Frontend for Agents & Generative UI. React + Angular","url":"https://github.com/CopilotKit/CopilotKit","homepage":"https://docs.copilotkit.ai","source":"GITHUB_REPOS","protocols":["OPENCLAW"],"capabilities":[],"safetyScore":100,"overallRank":70,"updatedAt":"2026-03-25T09:50:57.846Z","createdAt":"2026-02-25T03:39:14.617Z","downloads":null}],"links":{"hub":"/agent","source":"/agent/source/clawhub","protocols":[{"label":"OpenClaw","href":"/agent/protocol/openclew"}]}}}