Skip to main content

code-review-assistant

Analyzes PRs to assign expert reviewers, detect code issues, and generate architectural diagrams. Use for PR reviews, code analysis, pull requests, git diffs, or mentions of reviewers, code quality, or standards.

Datos de origen

Repositorio
pbakaus/code-review-assistant
Última actividad en el origen
24 de noviembre de 2025 a las 02:13
Idioma detectado de SKILL.md
inglés
Estrellas
10
Forks
4

Opciones de instalación

De forma predeterminada está seleccionado el prompt que primero revisa el origen. Puedes cambiar a un comando directo o descargar una copia local.

Revisa los archivos de origen

Lee SKILL.md y los archivos complementarios que muestra SkillsMP antes de decidir si quieres instalarlo.

Explorador de archivos
11 archivos

Mostrando SKILL.md

SKILL.md
Instrucciones de origen · Vista previa de solo lectura
name
code-review-assistant
description
Analyzes PRs to assign expert reviewers, detect code issues, and generate architectural diagrams. Use for PR reviews, code analysis, pull requests, git diffs, or mentions of reviewers, code quality, or standards.
allowed-tools
["Bash","Read","Write","Grep","Glob","WebFetch"]
# Code Review Assistant Analyzes a pull request to assign reviewers, detect code issues, and visually explain complex changes. Can automatically assign reviewers via GitHub CLI and post review comments. **Setup**: Run `bash scripts/setup.sh` to check environment and install dependencies. ## Get PR Data **If you have a diff and PR link provided:** Analyze that directly. **Otherwise:** Use GitHub CLI to fetch from current repo: ```bash gh pr view --json number,title,body,author,files gh pr diff ``` If `gh` not installed: https://cli.github.com/ ## Analysis 1. **Assign Reviewers**: Read `reference/team-expertise.md`, match expertise areas to changes in the diff 2. **Detect Issues**: - Read `reference/code-standards-map.md` for quick issue list - If any patterns match the diff, grep the full details from `code-standards.md` using the XML tag - Flag by severity (🔴🟡🔵) 3. **Visual Explanations**: Unless none or only minor code changes, run: - `node scripts/generate-diagram.js "<description>"` - `node scripts/upload-image.js <file>` → Cloudinary URL - Include as `![](url)` in review ### Output Format ```markdown ## 🤖 Claude PR Assessment ## 🎨 Visual Explanation ![](url) - Description (don't include if minimal code changes) ## 👥 Recommended Reviewers - **@username** (Name) - Expertise areas that match this PR ## ⚠️ Areas of Concern - 🔴/🟡/🔵 **Matched code standard pattern title** - Files it occurs in, optional 3-5 line snippets and aggregated explanation - Concise recommended solution ## ✅ Overall Assessment Strengths, changes needed, recommendation ``` ## Post-Review Actions After presenting the review, offer to help with: 1. **Auto-assign reviewers**: ```bash gh pr edit <number> --add-reviewer username1,username2,username3 ``` Use the GitHub usernames (@username) from the recommended reviewers list. 2. **Post review as PR comment**: ```bash gh pr comment <number> --body-file <review-file.md> ``` Save the full review markdown to a temporary file first, then post it. Ask the user if they'd like you to perform either of these actions. ## Reference Files Progressive disclosure - only load what you need: - `reference/team-expertise.md` - Team member expertise areas with GitHub usernames - `reference/code-standards-map.md` - Quick list of detectable issues (check this first) - `reference/code-standards.md` - Full issue details with XML tags (grep specific sections) ## Scripts - `scripts/generate-diagram.js` - Gemini API integration for visual explanations - `scripts/upload-image.js` - Cloudinary upload for public image URLs Run with Node.js, see script headers for usage details.
Ver en GitHub