codebase-agent.yml

name: Codebase Agent

on:
  issue_comment:
    types: [created]
  pull_request_review_comment:
    types: [created]
  issues:
    types: [opened, labeled]
  pull_request:
    types: [opened, labeled, ready_for_review]

permissions:
  contents: write
  pull-requests: write
  issues: write
  id-token: write  # Required for GCP Workload Identity

jobs:
  codebase-agent:
    runs-on: ubuntu-latest
    # Only run if:
    # 1. Comment contains @cba mention, OR
    # 2. Issue/PR has 'cba-review' label, OR
    # 3. Issue/PR has 'cba-help' label
    if: |
      (github.event_name == 'issue_comment' && contains(github.event.comment.body, '@cba')) ||
      (github.event_name == 'pull_request_review_comment' && contains(github.event.comment.body, '@cba')) ||
      contains(github.event.issue.labels.*.name, 'cba-review') ||
      contains(github.event.pull_request.labels.*.name, 'cba-review') ||
      contains(github.event.issue.labels.*.name, 'cba-help') ||
      contains(github.event.pull_request.labels.*.name, 'cba-help')

    steps:
      - name: Checkout repository
        uses: actions/checkout@v4
        with:
          fetch-depth: 0  # Full history for better context

      - name: Authenticate to Google Cloud
        id: auth
        uses: google-github-actions/auth@v2
        with:
          workload_identity_provider: ${{ secrets.GCP_WORKLOAD_IDENTITY_PROVIDER }}
          service_account: ${{ secrets.GCP_SERVICE_ACCOUNT }}

      - name: Set up Cloud SDK
        uses: google-github-actions/setup-gcloud@v2

      - name: Extract @cba command
        id: parse_command
        run: |
          # Parse the @cba mention and extract the command
          COMMENT_BODY="${{ github.event.comment.body }}"

          # Extract text after @cba mention
          if [[ "$COMMENT_BODY" =~ @cba[[:space:]]+(.*) ]]; then
            COMMAND="${BASH_REMATCH[1]}"
            echo "command=$COMMAND" >> $GITHUB_OUTPUT
            echo "has_mention=true" >> $GITHUB_OUTPUT
          else
            # No @cba mention, triggered by label
            echo "has_mention=false" >> $GITHUB_OUTPUT

            # Determine action based on label
            if [[ "${{ contains(github.event.issue.labels.*.name, 'cba-review') || contains(github.event.pull_request.labels.*.name, 'cba-review') }}" == "true" ]]; then
              echo "command=review this PR" >> $GITHUB_OUTPUT
            elif [[ "${{ contains(github.event.issue.labels.*.name, 'cba-help') || contains(github.event.pull_request.labels.*.name, 'cba-help') }}" == "true" ]]; then
              echo "command=help with this issue" >> $GITHUB_OUTPUT
            fi
          fi

      - name: Set up Python
        uses: actions/setup-python@v5
        with:
          python-version: '3.11'

      - name: Install dependencies
        run: |
          pip install google-cloud-aiplatform anthropic requests

      - name: Run Codebase Agent
        id: run_agent
        env:
          COMMAND: ${{ steps.parse_command.outputs.command }}
          HAS_MENTION: ${{ steps.parse_command.outputs.has_mention }}
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          ANTHROPIC_API_KEY: ${{ secrets.ANTHROPIC_API_KEY }}
          GCP_PROJECT_ID: ambient-prototypes
          GCP_REGION: us-central1
        run: |
          python3 << 'PYTHON_SCRIPT'
          import os
          import json
          import requests
          from anthropic import Anthropic

          # Get environment variables
          command = os.environ.get('COMMAND', 'help')
          github_token = os.environ['GITHUB_TOKEN']
          anthropic_api_key = os.environ['ANTHROPIC_API_KEY']

          # GitHub context
          github_context = json.loads('''${{ toJson(github) }}''')
          event_name = github_context['event_name']

          print(f"🤖 Codebase Agent activated!")
          print(f"Command: {command}")
          print(f"Event: {event_name}")

          # Prepare context for Claude
          repo_name = github_context['repository']

          if event_name in ['issue_comment', 'pull_request_review_comment']:
              issue_number = github_context['event']['issue']['number']
              issue_url = github_context['event']['issue']['html_url']
              context_type = "comment"
          elif 'pull_request' in github_context['event']:
              issue_number = github_context['event']['pull_request']['number']
              issue_url = github_context['event']['pull_request']['html_url']
              context_type = "pull_request"
          elif 'issue' in github_context['event']:
              issue_number = github_context['event']['issue']['number']
              issue_url = github_context['event']['issue']['html_url']
              context_type = "issue"

          # Call Claude API (using Anthropic directly for now)
          # In production, this would call Vertex AI using the GCP service account
          client = Anthropic(api_key=anthropic_api_key)

          prompt = f"""You are the Codebase Agent (CBA) for the {repo_name} repository.

You've been invoked with the command: {command}

Context:
- Type: {context_type}
- URL: {issue_url}
- Issue/PR: #{issue_number}

Please provide a helpful, concise response that addresses the user's request.
If this is a code review, focus on:
1. Code quality and best practices
2. Potential bugs or edge cases
3. Documentation and testing suggestions
4. Security considerations

If this is a help request, provide:
1. Clear explanation of the issue
2. Actionable next steps
3. Relevant documentation or examples"""

          message = client.messages.create(
              model="claude-3-5-sonnet-20241022",
              max_tokens=2000,
              messages=[{"role": "user", "content": prompt}]
          )

          response_text = message.content[0].text

          # Save response to file for next step
          with open('cba_response.txt', 'w') as f:
              f.write(response_text)

          print("✅ Codebase Agent response generated")
          PYTHON_SCRIPT

      - name: Post response to GitHub
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: |
          # Read the response from previous step
          RESPONSE=$(cat cba_response.txt)

          # Determine where to post the comment
          if [[ "${{ github.event_name }}" == "issue_comment" ]] || [[ "${{ github.event_name }}" == "pull_request_review_comment" ]]; then
            # Reply to the comment
            ISSUE_NUMBER="${{ github.event.issue.number }}"
          elif [[ "${{ github.event_name }}" == "pull_request" ]]; then
            ISSUE_NUMBER="${{ github.event.pull_request.number }}"
          elif [[ "${{ github.event_name }}" == "issues" ]]; then
            ISSUE_NUMBER="${{ github.event.issue.number }}"
          fi

          # Post comment using GitHub API
          curl -X POST \
            -H "Authorization: token $GITHUB_TOKEN" \
            -H "Accept: application/vnd.github.v3+json" \
            "https://api.github.com/repos/${{ github.repository }}/issues/$ISSUE_NUMBER/comments" \
            -d @- << EOF
          {
            "body": "## 🤖 Codebase Agent Response\n\n$RESPONSE\n\n---\n*Triggered by: ${{ github.event_name }} | Powered by Vertex AI*"
          }
          EOF

          echo "✅ Response posted to GitHub"

      - name: Add reaction to triggering comment
        if: github.event_name == 'issue_comment' || github.event_name == 'pull_request_review_comment'
        env:
          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
        run: |
          # Add 👀 reaction to show we're processing
          curl -X POST \
            -H "Authorization: token $GITHUB_TOKEN" \
            -H "Accept: application/vnd.github.v3+json" \
            "https://api.github.com/repos/${{ github.repository }}/issues/comments/${{ github.event.comment.id }}/reactions" \
            -d '{"content":"rocket"}'