Supercharge your GitHub workflow with AI-powered automation
AI GitHub Action leverages the OpenAI Agents framework to bring intelligent automation to your GitHub workflows. This action can analyze pull requests, issues, and code repositories to provide insightful feedback, automated code reviews, and helpful responses - all powered by OpenAI's advanced AI models.
- Automated PR Reviews - Get comprehensive code quality feedback with automatic PR approval for high-quality contributions
- Issue Analysis - Receive AI-generated responses to issues with categorization and prioritization
- Code Scanning - Identify security vulnerabilities and best practice violations across your codebase
- Custom Instructions - Tailor the AI behavior to focus on your project's specific needs
- Tracing - Powered by OpenAI's Agent SDK with built-in tracing through the OpenAI platform.
name: AI PR Review
on:
pull_request:
types: [opened, synchronize]
jobs:
pr-review:
runs-on: ubuntu-latest
permissions:
contents: read
pull-requests: write
steps:
- name: AI PR Review
uses: aguirreibarra/ai-github-action@main
with:
action-type: pr-review
openai-api-key: ${{ secrets.OPENAI_API_KEY }}
github-token: ${{ secrets.GITHUB_TOKEN }}
custom-prompt: "Focus on security best practices and performance optimization"The PR Review Action will analyze pull requests and post feedback by submiting a pull request review. The AI will provide specific feedback directly on code lines.
name: AI Issue Analysis
on:
issues:
types: [opened, labeled]
jobs:
analyze:
permissions:
issues: write
runs-on: ubuntu-latest
if: github.event.action == 'opened' || contains(github.event.issue.labels.*.name, 'needs-triage')
steps:
- name: AI Issue Analysis
uses: aguirreibarra/ai-github-action@main
with:
action-type: issue-analyze
openai-api-key: ${{ secrets.OPENAI_API_KEY }}
github-token: ${{ secrets.GITHUB_TOKEN }}
- name: Remove needs-triage label
uses: actions/github-script@v7
with:
github-token: ${{ secrets.GITHUB_TOKEN }}
script: |
try {
await github.rest.issues.removeLabel({
owner: context.repo.owner,
repo: context.repo.repo,
issue_number: context.issue.number,
name: 'needs-triage'
});
} catch (e) {
// Label might not exist, that's okay
}name: AI Code Scan
on:
push:
branches: [main]
schedule:
- cron: '0 0 * * 1' # Weekly scan
workflow_dispatch: # Allow manual runs
jobs:
scan:
permissions:
contents: read
issues: write
runs-on: ubuntu-latest
steps:
- name: AI Code Scan
uses: aguirreibarra/ai-github-action@main
with:
action-type: code-scan
openai-api-key: ${{ secrets.OPENAI_API_KEY }}
github-token: ${{ secrets.GITHUB_TOKEN }}| Input | Description | Required | Default |
|---|---|---|---|
action-type |
Type of action (pr-review, issue-analyze, code-scan) | Yes | - |
openai-api-key |
OpenAI API key | Yes | - |
github-token |
GitHub token for API access | Yes | - |
model |
OpenAI model to use | No | gpt-4o-mini |
max-turns |
Maximum turns for the AI | No | 30 |
custom-prompt |
Custom system prompt for the AI | No | - |
Control the verbosity of logs by setting the LOG_LEVEL environment variable:
steps:
- name: AI PR Review
uses: aguirreibarra/ai-github-action@main
with:
action-type: pr-review
openai-api-key: ${{ secrets.OPENAI_API_KEY }}
github-token: ${{ secrets.GITHUB_TOKEN }}
env:
LOG_LEVEL: DEBUG # Options: DEBUG, INFO, WARNING, ERROR, CRITICALWe welcome contributions! See CONTRIBUTING.md for guidelines.