These instructions apply to any AI agent working in the Chainguard Academy
docs repository. Codex, Cursor, and GitHub Copilot read this file directly.
Claude Code reads it through CLAUDE.md, and Gemini CLI through GEMINI.md.
This repository includes three writing skills in .agents/skills/. Claude
Code finds the same skills through symlinks in .claude/skills/. Apply them
before you produce any text, without waiting to be asked.
Use the writing-clearly-and-concisely skill for any prose a human will
read: documentation, pull request descriptions, commit messages, README
files, error messages, and code comments.
Use the written-style-guide skill for any content a Chainguard audience
will read, including docs, product UI strings, blog posts, and release
notes. It layers Chainguard's voice, tone, and product names on top of
writing-clearly-and-concisely, so load both. When reviewing someone's
text, flag issues and suggest fixes rather than rewriting it unless asked.
Its product names and their forms, including capitalization, take
precedence over every other guide.
Use the google-style skill when you create or edit technical
documentation. It adds the mechanical conventions of technical writing and
complements writing-clearly-and-concisely. When it conflicts with
written-style-guide on a docs page, follow google-style, except for
product names. Where the project's own
style guide
differs from Google's, follow the project guide.
Follow the Contributing section of README.md for page weights, images, tags, and pre-commit setup. A member of the docs team reviews every pull request, so flag anything you're unsure of in the pull request description rather than guessing.