proofread
Checks writing for style guide adherence and applies corrections.
Install
mkdir -p .claude/skills/proofread-ethersphere && curl -L -o skill.zip "https://agentskills.codes/api/skills/download/17390" && unzip -o skill.zip -d .claude/skills/proofread-ethersphere && rm skill.zipInstalls to .claude/skills/proofread-ethersphere
Activation
This is the description your AI agent reads to decide when to run this skill — the better it matches your request, the more reliably it fires.
Check text against Swarm voice and style guidelines, flag violations, and optionally apply fixesKey capabilities
- →Read text from a file path or inline input
- →Read Swarm voice and style guidelines
- →Identify violations of American English spelling
- →Flag incorrect usage of 'Swarm Foundation' vs 'Swarm'
- →Detect price or investment framing
- →Propose fixes for identified violations
How it works
The skill reads input text, compares it against the Swarm voice and style guidelines, identifies violations, and then presents a corrected version with a changelog.
Inputs & outputs
When to use proofread
- →Proofread project documentation
- →Check voice guidelines
- →Fix writing violations
About this skill
Proofread
Trigger: "Proofread: [text or file path]" or /proofread [text-or-file]
Purpose: Check writing against Swarm's voice and style guidelines, return a corrected version with the changes called out, and optionally apply the fixes to the original file.
Config keys used: None.
Step 1 — Get the input
From the argument:
- File path — read the file. Note that it's a file (you'll be able to apply fixes in Step 5).
- Inline text — work with the text directly. No file to write back to.
If the input is unclear or missing, ask. Don't proceed without real content.
Step 2 — Read the style guideline
Read style/Swarm-identity-guideline.md. It is the source of truth for voice, tone, and the non-negotiables — don't rely on a remembered list (per SKILL-STANDARD §1: discover, don't hardcode).
Step 3 — Check the text
Run the text through every rule in the guideline. The current non-negotiables include — but always cross-check the file:
- American English spelling
- "Swarm Foundation" for org actions; "Swarm" for the network
- No price or investment framing
- No overpromises — future items are goals/plans, not guarantees
- No contentious claims (e.g. "100% censorship-resistant")
- Concise — every sentence adds value
For each problem, capture: the exact phrase, which rule it violates, and the proposed fix.
Step 4 — Present the corrections
Show the user:
- The corrected version — the full text with fixes applied in place.
- A changelog — bullet list, one entry per change: what changed and which rule.
- Any violations you can't fix automatically (e.g. a contentious claim that needs rewriting from scratch) — flag them with a recommendation, don't silently rewrite.
Never fabricate corrections to make the text "punchier" — only flag real guideline violations.
Step 5 — Apply (optional)
If the input was a file, ask the user (AskUserQuestion): apply the corrections to the file now, or leave it?
- Apply — Edit the file with the corrected version.
- Leave — just present the corrections; the user copies what they want.
If the input was inline text, skip this step.
Step 6 — Verify
If fixes were applied, re-read the file and check:
- Every change in the changelog is present.
- No new typos or wording introduced beyond the listed changes.
Report: number of changes made, number of items flagged but not auto-fixed.
If the text is already clean, say so plainly. A 0-changes report is a fine outcome.
Notes
- Proofread is a leaf — no handoff. The user decides what to do with the corrections.
- Reference:
style/Swarm-identity-guideline.md.
When not to use it
- →When the input is unclear or missing
- →When the goal is to introduce new typos or wording beyond listed changes
- →When the goal is to make the text 'punchier' without guideline violations
Limitations
- →The skill only flags real guideline violations and does not fabricate corrections
- →The skill does not automatically fix violations that require rewriting from scratch
- →The skill requires `style/Swarm-identity-guideline.md` to be present
How it compares
This skill automates the application of specific organizational style guidelines and provides a detailed changelog, which is more precise than manual proofreading or general grammar checkers.
Compared to similar skills
proofread side by side with the closest alternatives in the catalog.
| Skill | Installs | Updated | Safety | Difficulty |
|---|---|---|---|---|
| proofread (this skill) | 0 | 1mo | No flags | Beginner |
| writing-clearly-and-concisely | 3 | 6mo | No flags | Beginner |
| professional-proofreader | 0 | 1mo | No flags | Beginner |
| writing-style | 0 | 1mo | No flags | Beginner |
Try saying
Example prompts that trigger this skill in your AI assistant.
More by ethersphere
View all by ethersphere →You might also like
writing-clearly-and-concisely
davila7
Use when writing prose humans will read—documentation, commit messages, error messages, explanations, reports, or UI text. Applies Strunk's timeless rules for clearer, stronger, more professional writing.
professional-proofreader
diegosouzapw
Professional Proofreader workflow skill. Use this skill when the user needs > and the operator should preserve the upstream workflow, copied support files, and provenance before merging or handing off.
writing-style
xpcoffee
Use when drafting any text. Specs, decisions, ADO/DevOps comments, code review comments, meeting notes, status updates, Teams messages, emails, PR descriptions, or any document. Applies hard style rules and a self-check before presenting.
write
get-convex
Writing style guide for technical content, social media, blog posts, READMEs, git commits, and developer documentation. Optimized to avoid AI detection patterns. Use when writing any content beyond code.
humanizer
davila7
Remove signs of AI-generated writing from text. Use when editing or reviewing text to make it sound more natural and human-written. Based on Wikipedia's comprehensive "Signs of AI writing" guide. Detects and fixes patterns including: inflated symbolism, promotional language, superficial -ing analyses, vague attributions, em dash overuse, rule of three, AI vocabulary words, negative parallelisms, and excessive conjunctive phrases. Credits: Original skill by @blader - https://github.com/blader/humanizer
humanizer-zh
op7418
去除文本中的 AI 生成痕迹。适用于编辑或审阅文本,使其听起来更自然、更像人类书写。 基于维基百科的"AI 写作特征"综合指南。检测并修复以下模式:夸大的象征意义、 宣传性语言、以 -ing 结尾的肤浅分析、模糊的归因、破折号过度使用、三段式法则、 AI 词汇、否定式排比、过多的连接性短语。