360 Secure SkillHub
360 安全技能中心

Humanizer Text Revision Helper

该 Skill 面向泛化、机械或与作者语气不符的草稿,先识别受众、渠道、事实和合规约束,再改写为更具体自然的文本,并输出编辑说明与风险提示;强调不用于规避检测器。

kyro-ma
kyro-ma
热度 16.07
最近更新 2026-07-13 20:56:20
分类:创作依赖:纯提示词应用场景:文案创作 / 校对润色

安装技能

复制下面的话术发给你的 Agent,它会通过 secure-skills 检索并安装该技能。

执行命令安装此 skill: npx secure-skills-cli add work-productivity-humanizer-remove-workflow-helper-040526
Metadata
name
work-productivity-humanizer-remove-workflow-helper
description
Revise AI-looking or over-polished text into clear, audience-appropriate writing while preserving truth, author intent, citations, and compliance constraints. Use when a user asks for humanizer, AI writing, editing, rewrite, voice, or needs practical workflow, code, checklist, documentation, or review support for this job.

Humanizer Text Revision Helper

Purpose

Use this skill when a draft sounds generic, robotic, repetitive, or misaligned with the author's voice, and the goal is ethical clarity, specificity, and audience fit.

Audience: writers, marketers, students, support teams, and skill authors who need natural text revision rather than evasive detector gaming.

Read references/requirement-plan.md when demand evidence, source links, scoring rationale, or review criteria are needed.

Workflow

  1. Identify the audience, channel, author voice, required facts, claims needing citations, forbidden wording, and desired tone.
  2. Mark symptoms of weak AI-style writing: vague claims, repeated rhythm, empty intensifiers, unsupported certainty, generic transitions, and unnatural politeness.
  3. Preserve factual content and citations first, then rewrite for concrete nouns, varied structure, direct verbs, and context-specific examples.
  4. Remove manipulative, deceptive, or detector-evasion framing; focus on readability and authenticity.
  5. Offer a minimal edit and a stronger rewrite when useful.
  6. Check the final text for factual drift, missing caveats, tone mismatch, and accidental changes to commitments or legal meaning.

Expected Outputs

  • A revised version of the user's text in the requested voice.
  • A short edit note explaining what changed and why.
  • Optional voice guidelines for future drafts.
  • A risk note for claims, citations, compliance, or sensitive context.

Validation

  • The rewrite keeps facts, intent, and required claims intact.
  • The text is more specific and readable without pretending to bypass detectors.
  • Tone matches audience and channel constraints.
  • Unsupported or risky claims are flagged rather than smoothed over.

Triggers

Keywords: humanizer, AI writing, editing, rewrite, voice, clarity, tone, copy review

Example trigger sentences:

  • Use $work-productivity-humanizer-remove-workflow-helper to make this draft sound more specific and natural.
  • Revise this AI-looking paragraph without changing the facts.
  • Create voice rules for rewriting support macros in our brand tone.