content-humanizer

分析一段内容或一个文件的"人味"（即像不像人类自然写出来的，AI 生成痕迹有多重），给出 0-100 量化打分，并在低于目标线（默认 80 分，用户可指定其他数值如90分）时在不改变原意的前提下改写到目标分数以上。触发条件：用户提到"人味"、"AI味"、"AI感"、"降AI率"、"去AI化"、"人性化改写"、"像人写的"、"过人工检测"、"过AI检测"、"让文章更自然"、"这段话是不是AI写的"、"帮我把这篇文章改得不像AI写的"，或者给出一段文字/文件并希望知道"读起来自然不自然"、"会不会被认出是AI写的"。即使用户只说"这段话感觉很机器，帮我改一下"或"帮我看看这篇稿子像不像真人写的"，也应使用本 skill。输出为 Markdown 报告，保存到当前项目 markdown/ 目录。

Quality

88%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

High

Do not use without reviewing

Quality

Content

85%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured, highly actionable skill body with a clear multi-step workflow, an explicit re-test feedback loop, and appropriate one-level-deep bundle references. Its only real weakness is verbosity in the design-rationale and threshold-philosophy prose.

Suggestions

Trim or remove the '为什么这么设计' rationale and the '这样做的好处是…' sentence — Claude only needs the two-layer methodology and weights, not the justification for them.

Shorten the threshold philosophy in step 4 to the rule (default 80, user-overridable, higher targets need more rounds) without the multi-sentence explanation of what the line 'really means'.

Move the design-philosophy framing out of the operational steps so the body leads with the workflow; keep the methodology description to the essential weights and layer split.

Dimension	Reasoning	Score
Conciseness	Mostly efficient operational prose, but the '这个 skill 在做什么，为什么这么设计' rationale, the '这样做的好处是…' justification, and the extended threshold-philosophy paragraph in step 4 are explanation Claude does not need and could be tightened.	3 / 5
Actionability	Fully executable: a copy-paste bash command for the scoring script (file and --text variants), an exact final-score formula, concrete before/after rewrite examples ('综上所述…' → direct conclusion), and a complete copy-paste Markdown report template.	5 / 5
Workflow Clarity	A clear 6-step sequence with an explicit validation feedback loop — after rewriting, re-run the script plus qualitative scoring, recompute, and retry up to 3 rounds — plus a '绝对不能做的事' constraint checklist for the destructive rewrite step.	5 / 5
Progressive Disclosure	The body is an overview; detailed qualitative anchors are split into the real one-level reference 'references/rubric.md' and statistical computation into 'scripts/human_score.py', both clearly signaled and present in the bundle.	5 / 5
	Total	18 / 20 Passed

Description

91%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong description: it states the skill's concrete actions, exposes a configurable threshold, and packs in an extensive, natural-sounding trigger list. It is mostly distinct from sibling skills, with only minor overlap on generic editing phrasing.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions — '给出 0-100 量化打分' and '在不改变原意的前提下改写到目标分数以上' with a configurable target line (default 80, user-specifiable) — but the skill is essentially score + rewrite, so coverage is good rather than comprehensive.	4 / 5
Completeness	Explicitly answers both 'what' (analyze humanness, score 0-100, rewrite above target without changing meaning) and 'when' via an explicit '触发条件：用户提到…' clause with concrete trigger phrases.	5 / 5
Trigger Term Quality	Comprehensive coverage of natural phrases users would say, including many synonyms and paraphrases — '人味', 'AI味', 'AI感', '降AI率', '去AI化', '过AI检测', '让文章更自然', '这段话是不是AI写的', '会不会被认出是AI写的', '感觉很机器帮我改一下'.	5 / 5
Distinctiveness Conflict Risk	Clear niche (AI-humanness detection/rewriting) with distinctive triggers like '降AI率'/'去AI化'/'过AI检测', but looser phrases such as '让文章更自然' or '帮我改一下' create minor overlap risk with general editing skills.	4 / 5
	Total	18 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: digoal/blog
Path: skills/skills_for_claude_web/content-humanizer/SKILL.md
Commit: 3b9c83d

Reviewed: 3 days ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.