Writing Guidelines
Audit docs and marketing prose against Vercel’s live writing handbook before you ship pages or publish changelog copy.
Overview
Writing Guidelines is an agent skill most often used in Ship (also Build, Launch) that fetches Vercel’s live writing handbook and audits your docs or prose for rule violations in file:line form.
Install
npx skills add https://github.com/vercel-labs/agent-skills --skill writing-guidelinesWhat is this skill?
- Fetches fresh rules from the public writing-guidelines command.md URL before every run
- Accepts a file or glob argument via frontmatter argument-hint
- Applies the full fetched rule set to each target file
- Reports violations in terse file:line format defined by the upstream guidelines
- Prompts for file scope when the user does not pass a path or pattern
Adoption & trust: 4.4k installs on skills.sh; 27.7k GitHub stars.
What problem does it solve?
You have draft docs or marketing copy but no consistent, up-to-date checklist to prove it matches your org’s voice, tone, and handbook rules.
Who is it for?
Indie builders maintaining product docs, changelogs, or landing copy who already follow (or want to adopt) Vercel-style writing rules and need agent-assisted audits on specific files.
Skip if: Teams that want generative rewriting from scratch, legal/compliance review, or reviews when you cannot fetch the guidelines URL (offline-only workflows).
When should I use this skill?
User asks to review docs, check writing style, audit prose, review docs voice and tone, or check a page against the writing handbook.
What do I get? / Deliverables
You get a line-level violation list aligned to the latest fetched guidelines so you can fix prose before merge or publish.
- Terse file:line finding list per upstream guideline output format
- Coverage of all rules from the freshly fetched handbook payload
Recommended Skills
Journey fit
Spans multiple journey phases - primary shelf plus alternate fits below.
Canonical shelf is Ship → Review because the skill is invoked as a quality gate on finished prose, similar to code review but for voice, tone, and handbook rules. Review subphase matches explicit triggers like “review my docs” and “audit prose” rather than initial authoring in Build → Docs.
Where it fits
Scan README and API reference sections for handbook violations before opening a docs PR.
Run a pre-release audit on changelog and error-message copy bundled with the release.
How it compares
Use as a handbook-backed prose linter, not as a generic grammar checker or unstructured “make this sound better” chat pass.
Common Questions / FAQ
Who is writing-guidelines for?
Solo and small-team builders who ship documentation and customer-facing prose and want agent-driven audits against Vercel’s published writing handbook.
When should I use writing-guidelines?
After you have draft markdown or copy—during Build → Docs cleanup, Ship → Review before release, or Launch → Content before publishing—when you are asked to review docs, check writing style, audit prose, or compare a page to the writing handbook.
Is writing-guidelines safe to install?
It instructs the agent to read your files and fetch a public GitHub raw URL; review the Security Audits panel on this Prism page and treat fetched remote content like any external dependency.
SKILL.md
READMESKILL.md - Writing Guidelines
# Writing Guidelines Review files for compliance with Writing Guidelines. ## How It Works 1. Fetch the latest guidelines from the source URL below 2. Read the specified files (or prompt user for files/pattern) 3. Check against all rules in the fetched guidelines 4. Output findings in the terse `file:line` format ## Guidelines Source Fetch fresh guidelines before each review: ``` https://raw.githubusercontent.com/vercel-labs/writing-guidelines/main/command.md ``` Use WebFetch to retrieve the latest rules. The fetched content contains all the rules and output format instructions. ## Usage When a user provides a file or pattern argument: 1. Fetch guidelines from the source URL above 2. Read the specified files 3. Apply all rules from the fetched guidelines 4. Output findings using the format specified in the guidelines If no files specified, ask the user which files to review.