Auto-Generate AI Knowledge with GitHub Actions

Keep your AI skills up-to-date automatically from any source with Skill Seekers v3.0.0 GitHub Action.

The Workflow

name: Update AI Skills
on:
  schedule:
    - cron: '0 0 * * 0'  # Weekly
  workflow_dispatch:  # Manual trigger

jobs:
  update-skills:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
      
      - name: Update React Skill from Docs
        uses: skill-seekers/action@v1
        with:
          config: configs/react.json
          format: langchain
          output: skills/react/
          
      - name: Update Internal API Skill from Repo
        run: |
          skill-seekers analyze --directory ./api --format langchain
          skill-seekers package output/api --target langchain
          
      - name: Commit changes
        run: |
          git config user.name "GitHub Action"
          git config user.email "action@github.com"
          git add skills/
          git commit -m "Update AI skills - $(date +%Y-%m-%d)"
          git push

Features

Multi-Source Support

Automatically process any source:

• 📚 Documentation - Auto-update when docs change
• 🐙 GitHub Repos - Track upstream framework changes
• 📄 PDF Files - Process updated manuals
• 💻 Local Codebases - Sync with your own code

Automated Updates

• ⏰ Scheduled runs - Weekly, daily, or custom schedule
• 🔄 On-demand - Manual trigger when needed
• 📝 Auto-commit - Changes committed automatically
• 📊 Notifications - Get notified of updates

Multi-Framework Support

strategy:
  matrix:
    source: 
      - { type: config, name: react }
      - { type: config, name: vue }
      - { type: github, url: https://github.com/django/django }
      
steps:
  - uses: skill-seekers/action@v1
    with:
      config: configs/${{ matrix.source.name }}.json
      format: langchain

Cloud Storage Integration

- uses: skill-seekers/action@v1
  with:
    config: configs/react.json
    format: langchain
    cloud: s3
    bucket: my-knowledge-base
  env:
    AWS_ACCESS_KEY_ID: ${{ secrets.AWS_KEY }}
    AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET }}

Complete Example

name: AI Knowledge Pipeline

on:
  schedule:
    - cron: '0 2 * * 0'  # Every Sunday at 2am
  workflow_dispatch:
  push:
    branches: [main]  # Also run when your codebase changes

jobs:
  build-skills:
    runs-on: ubuntu-latest
    strategy:
      matrix:
        include:
          # From documentation
          - source: docs
            config: react
            format: langchain
          # From GitHub
          - source: github
            repo: https://github.com/vuejs/vue
            format: llamaindex
          # From local codebase
          - source: local
            directory: ./internal-api
            format: markdown
    
    steps:
      - name: Checkout
        uses: actions/checkout@v4
        
      - name: Setup Python
        uses: actions/setup-python@v5
        with:
          python-version: '3.11'
          
      - name: Install Skill Seekers
        run: pip install skill-seekers
        
      - name: Generate Skill from Docs
        if: matrix.source == 'docs'
        run: |
          skill-seekers scrape \
            --config configs/${{ matrix.config }}.json \
            --format ${{ matrix.format }}
            
      - name: Generate Skill from GitHub
        if: matrix.source == 'github'
        run: |
          skill-seekers scrape \
            --format ${{ matrix.format }} \
            --github ${{ matrix.repo }}
            
      - name: Generate Skill from Local Code
        if: matrix.source == 'local'
        run: |
          skill-seekers analyze \
            --directory ${{ matrix.directory }} \
            --format ${{ matrix.format }}
            
      - name: Package Skill
        run: |
          skill-seekers package output/ --target ${{ matrix.format }}
            
      - name: Upload to Cloud
        run: |
          skill-seekers cloud upload output/ \
            --provider s3 \
            --bucket team-knowledge
        env:
          AWS_ACCESS_KEY_ID: ${{ secrets.AWS_KEY }}
          AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET }}

Docker Alternative

- name: Generate with Docker
  run: |
    docker run -v $(pwd):/data \
      skill-seekers:latest \
      scrape --config /data/config.json

Notifications

Get notified when skills update:

- name: Notify Slack
  if: success()
  uses: slackapi/slack-github-action@v1
  with:
    payload: |
      {
        "text": "🤖 AI Skills updated successfully!"
      }
  env:
    SLACK_WEBHOOK_URL: ${{ secrets.SLACK_WEBHOOK }}

Best Practices

1. Multi-source - Combine docs + repos + your own code
2. Schedule wisely - Don’t overwhelm docs servers
3. Use caching - Cache dependencies between runs
4. Monitor costs - Cloud storage has costs
5. Version control - Track skill changes in git
6. Test first - Run manually before scheduling

Result

Your AI knowledge stays fresh from all sources without manual work. Perfect for:

• Team knowledge bases (docs + internal code)
• Auto-updating documentation
• Tracking upstream framework changes
• CI/CD integrated workflows
• Multi-environment deployments

Start automating today!