| name | redbook-anti-risk-detoxifier-skills |
| description | Copywriting detoxification & compliance formatting skill. An essential compliance filter invoked prior to publishing content on domestic social media platforms (such as Redbook, TikTok, WeChat) to defend against image OCR surveillance and banned word detection. Crucial for neutralizing sensitive terms, hardcore tech jargon, and scraping rhetoric. |
🌟 Content Detoxifier & Compliance Formatter
Core Objective
Defend against platform OCR (Optical Character Recognition) mechanisms and text-based risk control systems. Ensure that the final images and body text generated for Redbook or other domestic social media do not trigger AI shadows, bans, or HTTP 451 errors due to high-risk keywords like "VPN", "Crypto", "Twitter scraping", "risk-control", or "reverse engineering".
Industry Standard Practices
For successful content matrixes and automated monetization pipelines, it is mandatory to bundle a Banned Word Scanner + Rewriting Engine. There are three core strategies:
- Implicit Replacement (Neutralization): Replace sensitive terminology with neutral commercial phrases.
- Positive Framing: Instead of talking about "defeating anti-bot systems, bypassing, or dark web", reframe the narrative around "underlying principles, super information arbitrage, and frontier insights".
- Erase Cross-Border Traces: DO NOT write "Twitter/X", "Telegram/TG", "Outside Network", "VPN", or "Overseas Platforms". Unify all such origins as "cutting-edge industry insights", "top-tier inner circles", or "first-hand internal information".
Operating Procedure
Step 1: Scan and Replace High-Risk Words
| Original High-Risk Term (BANNED) | Softened / Neutralized Alternative |
|---|
| Risk-control / Anti-ban / Avoid banning | Platform rules compliance / Account weight maintenance / Stability |
| Scraping / Spider / Reverse / Crawl | Automated collection / Efficiency tools / Automated data analysis |
| Bypass defenses / Login-free bypass | Protocol exploration / Browser automation black-tech |
| Break Great-Firewall / VPN / Proxy nodes | Magic network / Global vision / (Ideally, do not mention at all) |
| Crypto / Crypto-currency / Earn quick money | Web3 / Frontier tech / Digital assets / Commercial arbitrage |
| Twitter / X / Overseas network | Industry frontier insights / Top-tier institution circle / Internal intel |
Step 2: AIGC De-ranking Prevention and Authencity Boost
Redbook severely penalizes "low-quality AIGC" and "mass-produced AI text" (resulting in shadowbans or direct account bans).
- De-AI Labeling: Minimize direct usage of terms like "AI, Artificial Intelligence, Large Language Models, ChatGPT".
- Replacement Strategy: Replace "AI Butlers / AI Tools" with "Fully Automated Workflows, Cloud-based Digital Clones, Automation Assistants, Super External Brains, 24/7 Hosting Systems".
- Add Human Touch: You MUST provide practical, actionable value. Use first-person perspective actively ("I successfully ran this pipeline", "I tested it for 3 days", "I stepped on 5 traps"). Require emotional fluctuation and authentic user experience. Ban purely hollow lyrical sentences.
Step 3: Tone Softening
Rough Draft: "Teaching you how to write a Twitter spider to bypass risk-control and scrape Crypto data."
Post-Detox Version: "Sharing a super cool browser automation black-tech to effortlessly grab first-hand Web3 information arbitrage! 🚀"
Step 4: Output Self-Check (Mandatory)
- Do any high-risk words remain in the content destined for image rendering (Titles, Card text)?
- Does the overall tone sound like a "value-sharing lifestyle blogger" rather than a "black-hat hacker"?
HARD RULE: Before invoking redbook-content-creator-skills or render_xhs.py to generate ANY images or body text, the source draft MUST pass strictly through the scrubbing and rewriting procedures outlined above.
