用 Codex 或 Claude 帮你安装 复制这段 Prompt,粘贴到 Codex、Claude 或其他助手里,让它检查 Skill 页面并帮你完成安装。
直接命令不会经过审查 Prompt;运行前请先检查来源。
npx skills add https://github.com/tomevault-io/skills-registry --skill plain-language命令会保持在同一行。复制前请横向滚动并检查完整内容。
想先保存到本地?可下载 SkillsMP 当前能够提供的文件。
基于 SOC 职业分类
正在显示 SKILL.md
| name | plain-language |
| description | >- Use when this capability is needed. |
Review text files against the U.S. federal government Plain Language Guidelines (~30 rules across 6 categories) and produce a structured report of findings with concrete rewrites.
Important: This skill produces a report. Do not modify any reviewed files.
scripts/scan-plaintext-files.sh <project-directory> to discover all .md, .mdx, .markdown, .txt, .rtf, .rst, .adoc, .org, and .wiki files. The <project-directory> argument is required — it must be the root of the user's project (the repository being reviewed), NOT the skill's own directory.references/rules-quick-ref.md for the full rule checklist.references/severity-rubric.md to assign high/medium/low.plain-language-findings-YYYYMMDD.md in the project root (use today's date). Group findings by file, then by severity (high first). End with the summary block.Use this exact structure for each finding:
## [file path]
### Finding [N] — [Rule name] (severity: [high|medium|low])
- **Line [N]:** "[original text]"
- **Guideline:** [One-sentence explanation of the rule violated]
- **Suggested:** "[concrete rewrite]"
Example:
## docs/getting-started.md
### Finding 1 — Use simple words (severity: medium)
- **Line 14:** "In order to utilize the configuration module..."
- **Guideline:** Replace complex words with simple alternatives — "utilize" → "use", "in order to" → "to"
- **Suggested:** "To use the configuration module..."
### Finding 2 — Use active voice (severity: high)
- **Line 23:** "The database will be initialized by the setup script."
- **Guideline:** Make the actor the subject of the sentence
- **Suggested:** "The setup script initializes the database."
If a file has no findings, omit it from the report entirely — do not list it.
End the report with a summary:
## Summary
- **Files reviewed:** [N]
- **Total findings:** [N] ([N] high, [N] medium, [N] low)
- **Top issues:** [List the 2-3 most frequent rule violations]
Load references on demand to conserve context:
| File | When to load |
|---|---|
references/rules-quick-ref.md | Always — load at start of every review |
references/word-substitutions.md | When checking word choice or when you encounter a word that might have a simpler alternative |
references/active-voice-guide.md | When you detect passive voice patterns (forms of "to be" + past participle) |
references/before-and-after-examples.md | When crafting suggested rewrites — use as transformation models |
references/severity-rubric.md | When classifying findings — consult definitions and examples |
.md, .mdx, .markdown, .txt, .rtf, .rst, .adoc, .org, .wiki files; comments in source code files; UI strings; error messagesThe plain language guidelines and examples referenced by this skill originate from the U.S. federal government's Plain Language initiative, maintained by the General Services Administration (GSA). All guideline content is U.S. government work in the public domain. Source: https://digital.gov/guides/plain-language
Converted and distributed by TomeVault — claim your Tome and manage your conversions.