Skip to content

Installation

GitHub Marketplace

The easiest way to use this action is through the GitHub Actions marketplace:

  1. Visit the AI Workflow Failure Summary Action on GitHub Marketplace
  2. Click "Use latest version"
  3. Copy the usage snippet to your workflow

Manual Installation

- uses: ianlintner/ai_summary_action@v1.0.0
  with:
    github-token: ${{ secrets.GITHUB_TOKEN }}
    # ... other inputs

Using Latest Version

- uses: ianlintner/ai_summary_action@v1
  with:
    github-token: ${{ secrets.GITHUB_TOKEN }}
    # ... other inputs

- uses: ianlintner/ai_summary_action@main
  with:
    github-token: ${{ secrets.GITHUB_TOKEN }}
    # ... other inputs

Production Usage

Always pin to a specific version tag (like @v1.0.0) in production workflows to avoid unexpected breaking changes.

Prerequisites

1. GitHub Token

The action requires a GitHub token to access workflow logs and create issues. The default GITHUB_TOKEN is usually sufficient:

github-token: ${{ secrets.GITHUB_TOKEN }}

Required Permissions: 

permissions:
  actions: read       # To fetch workflow logs
  contents: read      # To access repository
  issues: write       # If creating issues (optional)

2. LLM Provider API Key

Choose one of the supported providers and obtain an API key:

Get API Key: 1. Visit OpenAI API Keys 2. Sign in or create an account 3. Create a new API key 4. Add to repository secrets as OPENAI_API_KEY

Pricing: Pay-as-you-go, starting at $0.03 per 1K tokens for GPT-4o-mini

Best For: General use, best balance of cost and quality

Get API Key: 1. No additional key needed! Use ${{ secrets.GITHUB_TOKEN }} 2. Visit GitHub Models Marketplace 3. Check available models and rate limits

Pricing: Free tier available with rate limits

Best For: Getting started, testing, small projects

Get API Key: 1. Visit Anthropic Console 2. Sign up for an account 3. Create an API key 4. Add to repository secrets as ANTHROPIC_API_KEY

Pricing: Pay-as-you-go, competitive with OpenAI

Best For: Advanced reasoning, large context windows

Get API Key: 1. Create an Azure OpenAI resource 2. Deploy a model (e.g., gpt-4o) 3. Get the endpoint and API key 4. Add to secrets: - AZURE_OPENAI_API_KEY - AZURE_OPENAI_ENDPOINT - AZURE_OPENAI_DEPLOYMENT

Pricing: Enterprise pricing through Azure

Best For: Enterprise deployments, compliance requirements

Adding Secrets to Your Repository

  1. Go to your repository on GitHub
  2. Click Settings → Secrets and variables → Actions
  3. Click New repository secret
  4. Add each required secret:
    • Name: OPENAI_API_KEY (or your provider)
    • Value: Your API key
  5. Click Add secret

Organization Secrets

For multiple repositories:

  1. Go to your organization on GitHub
  2. Click Settings → Secrets and variables → Actions
  3. Click New organization secret
  4. Add the secret and select which repositories can access it

Environment Secrets

For environment-specific configuration:

  1. Create an environment in your repository settings
  2. Add secrets to that environment
  3. Reference in your workflow:
jobs:
  analyze:
    runs-on: ubuntu-latest
    environment: production  # Use environment secrets
    steps:
      - uses: ianlintner/ai_summary_action@v1
        with:
          openai-api-key: ${{ secrets.OPENAI_API_KEY }}

Verification

To verify installation, add a simple test workflow:

.github/workflows/test-ai-action.yml
name: Test AI Action

on:
  workflow_dispatch:  # Manual trigger

jobs:
  fail-intentionally:
    runs-on: ubuntu-latest
    steps:
      - name: This will fail
        run: exit 1

  analyze:
    runs-on: ubuntu-latest
    if: failure()
    needs: [fail-intentionally]
    steps:
      - name: Analyze Failure
        uses: ianlintner/ai_summary_action@v1
        with:
          github-token: ${{ secrets.GITHUB_TOKEN }}
          llm-provider: 'openai'
          openai-api-key: ${{ secrets.OPENAI_API_KEY }}

Run this workflow manually from the Actions tab to test your setup.

System Requirements

  • GitHub Actions: Any runner (ubuntu-latest, windows-latest, macos-latest)
  • Node.js Runtime: Node 20 (provided by GitHub Actions)
  • Internet Access: Required to call LLM provider APIs
  • Permissions: Read access to Actions, optional write access to Issues

Upgrading

To upgrade to a newer version:

  1. Check the Changelog for breaking changes
  2. Update the version in your workflow: 
    - uses: ianlintner/ai_summary_action@v2.0.0  # New version
  3. Test in a non-production workflow first
  4. Roll out to all workflows

Uninstalling

To remove the action:

  1. Delete or comment out the action step in your workflows
  2. Optionally remove the API key secret from repository settings
  3. Remove any custom prompt files from .github/prompts/

Next Steps