michelangelo

Trigger when implementing any new feature, hook, or component — decisions about where to put new files and how to import them must follow these rules. Also trigger when adding imports.

Trigger when implementing, designing, or modifying any React component or hook — including planning new components before any file is touched. Enforces styling and naming conventions that OVERRIDE default React patterns.

Trigger when testing is any part of the task — writing tests, adding coverage, adding test cases, or implementing code that will need tests. Project-specific mocking rules that OVERRIDE standard defaults.

Create or update a GitHub PR using the repo's pull_request_template.md format. Use when creating a new PR or updating an existing PR body.

Update Michelangelo documentation. Use when adding, modifying, or fixing docs in the docs/ folder. Ensures consistent formatting, proper titles, and valid links.

Create a 3-agent team (engineer, tech-writer, product-manager) to review and improve docs for open-source quality. Use when reviewing any docs/ directory or specific doc files.

Trigger when implementing anything in .ts or .tsx — new types, interfaces, hooks, or any feature that will involve TypeScript decisions. Don't wait for type errors; consult this before writing types, not after. Covers patterns that OVERRIDE TypeScript defaults.

Apply when adding, writing, or reviewing comments or JSDoc in any file. Also apply when making significant code changes that introduce non-obvious logic, public APIs, or edge cases that likely need documentation. Covers when to write JSDoc vs. when to skip it entirely, and inline comment standards.

Trigger whenever any code will be written or changed — including features, refactors, and test changes. If the task ends with modified files, this skill applies. Skip only for pure read-only questions like explaining or debugging without changes.

Apply Python best practices, conventions, and style rules from Google's Python Style Guide. Use when writing, reviewing, or refactoring Python code to ensure clean, maintainable, and Pythonic implementations.