Hiddenden/openrabbit
3
0
Fork 0
Code Issues Pull Requests Actions Packages Projects 1 Releases Wiki Activity
Files
95937c4738c714df8241d70dbd69f20ad6e02b72
openrabbit/docs/workflows.md
latte a1fe47cdf4 update
2026-01-07 20:21:33 +01:00

22 KiB
Raw Blame History

Workflows

This document provides ready-to-use workflow files for integrating OpenRabbit AI code review into your Gitea or GitHub repositories.

Simply copy the workflow files below into your repository's .gitea/workflows/ or .github/workflows/ directory.

Quick Start

  1. Copy the workflow files to your repository
  2. Configure the required secrets (see Required Secrets)
  3. Update the placeholders:
    • YourOrg/OpenRabbit - Your OpenRabbit fork repository
    • https://git.example.com/api/v1 - Your Gitea instance API URL
    • Bartender - Your bot's username (for self-trigger prevention)

Gitea Workflows

1. PR Review Workflow

Automatically reviews pull requests when opened or updated.

# .gitea/workflows/enterprise-ai-review.yml
name: Enterprise AI Code Review

on:
  pull_request:
    types: [opened, synchronize]

jobs:
  ai-review:
    runs-on: ubuntu-latest

    steps:
      # Checkout the PR repository
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0

      # Checkout the CENTRAL AI tooling repo
      - uses: actions/checkout@v4
        with:
          repository: YourOrg/OpenRabbit
          path: .ai-review
          token: ${{ secrets.AI_REVIEW_TOKEN }}

      # Setup Python
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"

      # Install dependencies
      - run: pip install requests pyyaml

      # Run the AI review
      - name: Run Enterprise AI Review
        env:
          AI_REVIEW_TOKEN: ${{ secrets.AI_REVIEW_TOKEN }}
          AI_REVIEW_REPO: ${{ gitea.repository }}
          AI_REVIEW_API_URL: https://git.example.com/api/v1
          AI_REVIEW_PR_NUMBER: ${{ gitea.event.pull_request.number }}

          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
          OPENROUTER_API_KEY: ${{ secrets.OPENROUTER_API_KEY }}
          OLLAMA_HOST: ${{ secrets.OLLAMA_HOST }}
        run: |
          cd .ai-review/tools/ai-review
          python main.py pr ${{ gitea.repository }} ${{ gitea.event.pull_request.number }} \
            --title "${{ gitea.event.pull_request.title }}"

      # Fail CI on HIGH severity (optional)
      - name: Check Review Result
        if: failure()
        run: |
          echo "AI Review found HIGH severity issues. Please address them before merging."
          exit 1

2. Issue Triage Workflow

Triggers on @codebot triage command in issue comments.

# .gitea/workflows/ai-issue-triage.yml
name: AI Issue Triage

# WORKFLOW ROUTING:
# This workflow handles ONLY the 'triage' command
# Other workflows: ai-comment-reply.yml (specific commands), ai-chat.yml (free-form questions)

on:
    issue_comment:
        types: [created]

jobs:
    ai-triage:
        runs-on: ubuntu-latest
        # Only run if comment contains @codebot triage
        # CRITICAL: Ignore bot's own comments to prevent infinite loops (bot username: Bartender)
        if: |
            github.event.comment.user.login != 'Bartender' &&
            contains(github.event.comment.body, '@codebot triage')
        steps:
            - uses: actions/checkout@v4

            - uses: actions/checkout@v4
              with:
                  repository: YourOrg/OpenRabbit
                  path: .ai-review
                  token: ${{ secrets.AI_REVIEW_TOKEN }}

            - uses: actions/setup-python@v5
              with:
                  python-version: "3.11"

            - run: pip install requests pyyaml

            - name: Run AI Issue Triage
              env:
                  AI_REVIEW_TOKEN: ${{ secrets.AI_REVIEW_TOKEN }}
                  AI_REVIEW_REPO: ${{ gitea.repository }}
                  AI_REVIEW_API_URL: https://git.example.com/api/v1
                  OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
                  OPENROUTER_API_KEY: ${{ secrets.OPENROUTER_API_KEY }}
                  OLLAMA_HOST: ${{ secrets.OLLAMA_HOST }}
              run: |
                  cd .ai-review/tools/ai-review
                  python main.py issue ${{ gitea.repository }} ${{ gitea.event.issue.number }}

3. Comment Reply Workflow

Handles specific commands: help, explain, suggest, security, summarize, changelog, explain-diff, review-again, setup-labels.

# .gitea/workflows/ai-comment-reply.yml
name: AI Comment Reply

# WORKFLOW ROUTING:
# This workflow handles SPECIFIC commands: help, explain, suggest, security, summarize, changelog, explain-diff, review-again, setup-labels
# Other workflows: ai-issue-triage.yml (@codebot triage), ai-chat.yml (free-form questions)

on:
    issue_comment:
        types: [created]

# CUSTOMIZE YOUR BOT NAME:
# Change '@codebot' in the 'if' condition below to match your config.yml mention_prefix
# Examples: '@bartender', '@uni', '@joey', '@codebot'

jobs:
    ai-reply:
        runs-on: ubuntu-latest
        # Only run for specific commands (not free-form chat or triage)
        # This prevents duplicate runs with ai-chat.yml and ai-issue-triage.yml
        # CRITICAL: Ignore bot's own comments to prevent infinite loops (bot username: Bartender)
        if: |
            github.event.comment.user.login != 'Bartender' &&
            (contains(github.event.comment.body, '@codebot help') ||
             contains(github.event.comment.body, '@codebot explain') ||
             contains(github.event.comment.body, '@codebot suggest') ||
             contains(github.event.comment.body, '@codebot security') ||
             contains(github.event.comment.body, '@codebot summarize') ||
             contains(github.event.comment.body, '@codebot changelog') ||
             contains(github.event.comment.body, '@codebot explain-diff') ||
             contains(github.event.comment.body, '@codebot review-again') ||
             contains(github.event.comment.body, '@codebot setup-labels'))
        steps:
            - uses: actions/checkout@v4

            - uses: actions/checkout@v4
              with:
                  repository: YourOrg/OpenRabbit
                  path: .ai-review
                  token: ${{ secrets.AI_REVIEW_TOKEN }}

            - uses: actions/setup-python@v5
              with:
                  python-version: "3.11"

            - run: pip install requests pyyaml

            - name: Run AI Comment Response
              env:
                  AI_REVIEW_TOKEN: ${{ secrets.AI_REVIEW_TOKEN }}
                  AI_REVIEW_API_URL: https://git.example.com/api/v1
                  OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
                  OPENROUTER_API_KEY: ${{ secrets.OPENROUTER_API_KEY }}
                  OLLAMA_HOST: ${{ secrets.OLLAMA_HOST }}
              run: |
                  cd .ai-review/tools/ai-review

                  # Determine if this is a PR or issue comment
                  IS_PR="${{ gitea.event.issue.pull_request != null }}"
                  REPO="${{ gitea.repository }}"
                  ISSUE_NUMBER="${{ gitea.event.issue.number }}"

                  # Validate inputs
                  if [ -z "$REPO" ] || [ -z "$ISSUE_NUMBER" ]; then
                      echo "Error: Missing required parameters"
                      exit 1
                  fi

                  # Validate repository format (owner/repo)
                  if ! echo "$REPO" | grep -qE '^[a-zA-Z0-9_-]+/[a-zA-Z0-9_-]+$'; then
                      echo "Error: Invalid repository format: $REPO"
                      exit 1
                  fi

                  if [ "$IS_PR" = "true" ]; then
                      # This is a PR comment - use safe dispatch with minimal event data
                      # Build minimal event payload (does not include sensitive user data)
                      EVENT_DATA=$(cat <<EOF
                  {
                      "action": "created",
                      "issue": {
                          "number": ${{ gitea.event.issue.number }},
                          "pull_request": {}
                      },
                      "comment": {
                          "id": ${{ gitea.event.comment.id }},
                          "body": $(echo '${{ gitea.event.comment.body }}' | jq -Rs .)
                      }
                  }
                  EOF
                  )

                      # Use safe dispatch utility
                      python utils/safe_dispatch.py issue_comment "$REPO" "$EVENT_DATA"
                  else
                      # This is an issue comment - use the comment command
                      COMMENT_BODY='${{ gitea.event.comment.body }}'
                      python main.py comment "$REPO" "$ISSUE_NUMBER" "$COMMENT_BODY"
                  fi

4. Chat Workflow (Free-form Questions)

Handles any @codebot mention that isn't a specific command - for asking questions about the codebase.

# .gitea/workflows/ai-chat.yml
name: AI Chat (Bartender)

# WORKFLOW ROUTING:
# This workflow handles FREE-FORM questions/chat (no specific command)
# Other workflows: ai-issue-triage.yml (@codebot triage), ai-comment-reply.yml (specific commands)
# This is the FALLBACK for any @codebot mention that isn't a known command

on:
    issue_comment:
        types: [created]

# CUSTOMIZE YOUR BOT NAME:
# Change '@codebot' in all conditions below to match your config.yml mention_prefix
# Examples: '@bartender', '@uni', '@joey', '@codebot'

jobs:
    ai-chat:
        # Only run if comment mentions the bot but NOT a specific command
        # This prevents duplicate runs with ai-comment-reply.yml and ai-issue-triage.yml
        # CRITICAL: Ignore bot's own comments to prevent infinite loops (bot username: Bartender)
        if: |
            github.event.comment.user.login != 'Bartender' &&
            contains(github.event.comment.body, '@codebot') &&
            !contains(github.event.comment.body, '@codebot triage') &&
            !contains(github.event.comment.body, '@codebot help') &&
            !contains(github.event.comment.body, '@codebot explain') &&
            !contains(github.event.comment.body, '@codebot suggest') &&
            !contains(github.event.comment.body, '@codebot security') &&
            !contains(github.event.comment.body, '@codebot summarize') &&
            !contains(github.event.comment.body, '@codebot changelog') &&
            !contains(github.event.comment.body, '@codebot explain-diff') &&
            !contains(github.event.comment.body, '@codebot review-again') &&
            !contains(github.event.comment.body, '@codebot setup-labels')
        runs-on: ubuntu-latest
        steps:
            - uses: actions/checkout@v4

            - uses: actions/checkout@v4
              with:
                  repository: YourOrg/OpenRabbit
                  path: .ai-review
                  token: ${{ secrets.AI_REVIEW_TOKEN }}

            - uses: actions/setup-python@v5
              with:
                  python-version: "3.11"

            - run: pip install requests pyyaml

            - name: Run AI Chat
              env:
                  AI_REVIEW_TOKEN: ${{ secrets.AI_REVIEW_TOKEN }}
                  AI_REVIEW_REPO: ${{ gitea.repository }}
                  AI_REVIEW_API_URL: https://git.example.com/api/v1
                  OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
                  OPENROUTER_API_KEY: ${{ secrets.OPENROUTER_API_KEY }}
                  OLLAMA_HOST: ${{ secrets.OLLAMA_HOST }}
                  SEARXNG_URL: ${{ secrets.SEARXNG_URL }}
              run: |
                  cd .ai-review/tools/ai-review
                  python main.py comment ${{ gitea.repository }} ${{ gitea.event.issue.number }} "${{ gitea.event.comment.body }}"

5. Codebase Analysis Workflow

Weekly scheduled codebase health analysis.

# .gitea/workflows/ai-codebase-review.yml
name: AI Codebase Quality Review

on:
  # Weekly scheduled run
  schedule:
    - cron: "0 0 * * 0" # Every Sunday at midnight

  # Manual trigger
  workflow_dispatch:
    inputs:
      report_type:
        description: "Type of report to generate"
        required: false
        default: "full"
        type: choice
        options:
          - full
          - security
          - quick

jobs:
  ai-codebase-review:
    runs-on: ubuntu-latest

    steps:
      # Checkout the repository
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0 # Full history for analysis

      # Checkout central AI tooling
      - uses: actions/checkout@v4
        with:
          repository: YourOrg/OpenRabbit
          path: .ai-review
          token: ${{ secrets.AI_REVIEW_TOKEN }}

      # Setup Python
      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"

      # Install dependencies
      - run: pip install requests pyyaml

      # Run AI codebase analysis
      - name: Run AI Codebase Analysis
        env:
          AI_REVIEW_TOKEN: ${{ secrets.AI_REVIEW_TOKEN }}
          AI_REVIEW_REPO: ${{ gitea.repository }}
          AI_REVIEW_API_URL: https://git.example.com/api/v1

          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
          OPENROUTER_API_KEY: ${{ secrets.OPENROUTER_API_KEY }}
          OLLAMA_HOST: ${{ secrets.OLLAMA_HOST }}
        run: |
          cd .ai-review/tools/ai-review
          python main.py codebase ${{ gitea.repository }}

GitHub Workflows

For GitHub, use ${{ github.* }} instead of ${{ gitea.* }} and ${{ secrets.GITHUB_TOKEN }} for authentication.

1. PR Review Workflow (GitHub)

# .github/workflows/ai-review.yml
name: AI Code Review

on:
  pull_request:
    types: [opened, synchronize]

jobs:
  ai-review:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0

      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"

      - run: pip install requests pyyaml

      - name: Run AI Review
        env:
          AI_REVIEW_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          AI_REVIEW_REPO: ${{ github.repository }}
          AI_REVIEW_API_URL: https://api.github.com
          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
        run: |
          cd tools/ai-review
          python main.py pr ${{ github.repository }} ${{ github.event.pull_request.number }}

2. Issue Triage Workflow (GitHub)

# .github/workflows/ai-issue-triage.yml
name: AI Issue Triage

on:
  issue_comment:
    types: [created]

jobs:
  ai-triage:
    runs-on: ubuntu-latest
    # CRITICAL: Update 'YourBotUsername' to your bot's actual GitHub username
    if: |
      github.event.comment.user.login != 'YourBotUsername' &&
      contains(github.event.comment.body, '@codebot triage')
    steps:
      - uses: actions/checkout@v4

      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"

      - run: pip install requests pyyaml

      - name: Run AI Issue Triage
        env:
          AI_REVIEW_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          AI_REVIEW_REPO: ${{ github.repository }}
          AI_REVIEW_API_URL: https://api.github.com
          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
        run: |
          cd tools/ai-review
          python main.py issue ${{ github.repository }} ${{ github.event.issue.number }}

3. Comment Reply Workflow (GitHub)

# .github/workflows/ai-comment-reply.yml
name: AI Comment Reply

on:
  issue_comment:
    types: [created]

jobs:
  ai-reply:
    runs-on: ubuntu-latest
    # CRITICAL: Update 'YourBotUsername' to your bot's actual GitHub username
    if: |
      github.event.comment.user.login != 'YourBotUsername' &&
      (contains(github.event.comment.body, '@codebot help') ||
       contains(github.event.comment.body, '@codebot explain') ||
       contains(github.event.comment.body, '@codebot suggest') ||
       contains(github.event.comment.body, '@codebot security') ||
       contains(github.event.comment.body, '@codebot summarize') ||
       contains(github.event.comment.body, '@codebot changelog') ||
       contains(github.event.comment.body, '@codebot explain-diff') ||
       contains(github.event.comment.body, '@codebot review-again') ||
       contains(github.event.comment.body, '@codebot setup-labels'))
    steps:
      - uses: actions/checkout@v4

      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"

      - run: pip install requests pyyaml

      - name: Run AI Comment Response
        env:
          AI_REVIEW_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          AI_REVIEW_REPO: ${{ github.repository }}
          AI_REVIEW_API_URL: https://api.github.com
          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
          SEARXNG_URL: ${{ secrets.SEARXNG_URL }}
        run: |
          cd tools/ai-review
          python main.py comment ${{ github.repository }} ${{ github.event.issue.number }} "${{ github.event.comment.body }}"

4. Chat Workflow (GitHub)

# .github/workflows/ai-chat.yml
name: AI Chat

on:
  issue_comment:
    types: [created]

jobs:
  ai-chat:
    runs-on: ubuntu-latest
    # CRITICAL: Update 'YourBotUsername' to your bot's actual GitHub username
    if: |
      github.event.comment.user.login != 'YourBotUsername' &&
      contains(github.event.comment.body, '@codebot') &&
      !contains(github.event.comment.body, '@codebot triage') &&
      !contains(github.event.comment.body, '@codebot help') &&
      !contains(github.event.comment.body, '@codebot explain') &&
      !contains(github.event.comment.body, '@codebot suggest') &&
      !contains(github.event.comment.body, '@codebot security') &&
      !contains(github.event.comment.body, '@codebot summarize') &&
      !contains(github.event.comment.body, '@codebot changelog') &&
      !contains(github.event.comment.body, '@codebot explain-diff') &&
      !contains(github.event.comment.body, '@codebot review-again') &&
      !contains(github.event.comment.body, '@codebot setup-labels')
    steps:
      - uses: actions/checkout@v4

      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"

      - run: pip install requests pyyaml

      - name: Run AI Chat
        env:
          AI_REVIEW_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          AI_REVIEW_REPO: ${{ github.repository }}
          AI_REVIEW_API_URL: https://api.github.com
          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
          SEARXNG_URL: ${{ secrets.SEARXNG_URL }}
        run: |
          cd tools/ai-review
          python main.py comment ${{ github.repository }} ${{ github.event.issue.number }} "${{ github.event.comment.body }}"

5. Codebase Analysis Workflow (GitHub)

# .github/workflows/ai-codebase-review.yml
name: AI Codebase Analysis

on:
  schedule:
    - cron: "0 0 * * 0"  # Weekly on Sunday
  workflow_dispatch:

jobs:
  ai-codebase:
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
        with:
          fetch-depth: 0

      - uses: actions/setup-python@v5
        with:
          python-version: "3.11"

      - run: pip install requests pyyaml

      - name: Run Codebase Analysis
        env:
          AI_REVIEW_TOKEN: ${{ secrets.GITHUB_TOKEN }}
          AI_REVIEW_REPO: ${{ github.repository }}
          AI_REVIEW_API_URL: https://api.github.com
          OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
        run: |
          cd tools/ai-review
          python main.py codebase ${{ github.repository }}

Required Secrets

Gitea

Secret Required Description
AI_REVIEW_TOKEN Yes Gitea bot access token with repo read/write
OPENAI_API_KEY Choose one OpenAI API key
OPENROUTER_API_KEY Choose one OpenRouter API key
OLLAMA_HOST Choose one Ollama server URL (e.g., http://localhost:11434)
SEARXNG_URL Optional SearXNG instance for web search

GitHub

Secret Required Description
GITHUB_TOKEN Auto Built-in token (automatic)
OPENAI_API_KEY Choose one OpenAI API key
OPENROUTER_API_KEY Choose one OpenRouter API key
OLLAMA_HOST Choose one Ollama server URL
SEARXNG_URL Optional SearXNG instance for web search

Customization Guide

Changing the Bot Name

  1. Update config.yml:

    interaction:
  mention_prefix: "@yourbot"

  2. Update all workflow files - replace @codebot with your bot name

  3. Update the bot username check - replace 'Bartender' with your bot's username

Using a Different LLM Provider

Set the appropriate secret and update config.yml:

# For OpenRouter (Claude, etc.)
provider: openrouter
model:
  openrouter: anthropic/claude-3.5-sonnet

# For Ollama (self-hosted)
provider: ollama
model:
  ollama: codellama:13b

Disabling Specific Features

In config.yml:

agents:
  pr:
    enabled: true
    security_scan: true      # Set to false to disable security scanning
    inline_comments: true    # Set to false to disable inline comments
    auto_summary:
      enabled: true          # Set to false to disable auto-summary for empty PRs
  
  issue:
    enabled: true
    auto_label: true         # Set to false to disable auto-labeling

Workflow Routing Logic

The workflows are designed to be mutually exclusive to prevent duplicate runs:

@codebot mention detected
├── Contains "triage"?
│   └── YES → ai-issue-triage.yml
├── Contains specific command (help, explain, suggest, etc.)?
│   └── YES → ai-comment-reply.yml
└── Just @codebot + question (no command)?
    └── YES → ai-chat.yml

Important: All workflows include a check to ignore the bot's own comments, preventing infinite loops.

Troubleshooting

Duplicate Workflow Runs

If you see multiple workflows triggering:

  1. Check that command conditions in ai-chat.yml exclude all commands from ai-comment-reply.yml
  2. Verify the bot username check is correct

Bot Not Responding

  1. Check that secrets are configured correctly
  2. Verify the API URL is correct for your platform
  3. Check workflow logs for errors

Infinite Loop Prevention

All workflows include:

if: github.event.comment.user.login != 'BotUsername'

Make sure to update 'BotUsername' to your actual bot's login name.

Example Usage

After setting up the workflows:

# Triage an issue
@codebot triage

# Get help
@codebot help

# Ask a question about the codebase
@codebot how does authentication work?

# Generate PR summary
@codebot summarize

# Generate changelog
@codebot changelog

# Re-run PR review
@codebot review-again

# Setup repository labels
@codebot setup-labels
Reference in New Issue View Git Blame Copy Permalink