AI Issue Resolver - using Langchain - By AxioTree

🤖 A GitHub Action plugin that leverages AI-powered code generation and pull request management using Langchain SDK.

Features

• 🎯 Automated PR Generation

  • Generate PRs automatically from issues tagged with ai-issue-resolver-pr
  • AI analysis of issue descriptions for accurate code changes
  • Intelligent branch management and commit organization

• 🔍 Advanced Code Review

  • Comprehensive code quality analysis
  • Performance optimization suggestions
  • Design pattern recommendations
  • Testing coverage assessment
  • Security vulnerability scanning

• 🛡️ Security Analysis

  • Input validation vulnerability detection
  • Authentication/authorization issue identification
  • Data exposure risk assessment
  • Injection vulnerability scanning
  • Dependency security audit

• 💬 Interactive PR Updates

  • Update PRs using slash commands
  • Request specific changes with natural language
  • Get instant AI-powered feedback
  • Multiple command support:
    • /ai-issue-resolver-change - Request specific code changes
    • /ai-issue-resolver-review - Trigger comprehensive code review

Installation

From GitHub Marketplace (Recommended)

1. Visit the Axiotree AI Issue Resolver in the GitHub Marketplace
2. Click "Install it for free"
3. Configure the action with required secrets
4. Add the workflow file to your repository

Quick Setup

1. Create .github/workflows/ai-pr.yml in your repository:

name: AI PR Generation
on:
  issues:
    types: [labeled]
  issue_comment:
    types: [created]

jobs:
  ai-pr:
    runs-on: ubuntu-latest
    if: |
      (github.event_name == 'issues' && contains(github.event.issue.labels.*.name, 'ai-issue-resolver-pr')) ||
      (github.event_name == 'issue_comment' && startsWith(github.event.comment.body, '/ai-issue-resolver'))
    steps:
      - uses: actions/checkout@v3
      - uses: Axiotree/ai-issue-resolver
        with:
          github-token: ${{ secrets.GITHUB_TOKEN }}
          model-api-key: ${{ secrets.MODEL_API_KEY }}
          model-provider: 'openai'  # optional
          model-name: 'gpt-4'      # optional
          base-branch: 'main'      # optional
          review-threshold: '0.8'   # optional

2. Add required secrets in your repository settings:
   • GITHUB_TOKEN: Automatically provided by GitHub
   • MODEL_API_KEY: Your Language Model API key (supports OpenAI, Anthropic, etc.)

Usage

1. Generating AI PRs from Issues

1. Create a new issue describing the desired changes
2. Add the ai-issue-resolver-pr label
3. The action will automatically:
   • Analyze the issue using AI
   • Generate code changes
   • Create a PR with the changes
   • Add relevant labels

Example issue:

Title: Add input validation to user registration

Description:
Please add input validation for:
- Email format checking
- Password strength requirements (min 8 chars, special chars)
- Username validation (alphanumeric only)

2. Using Code Review

Comment on any PR with:

/ai-issue-resolver-review

The AI will provide:

• Code quality assessment
• Security scan results
• Performance suggestions
• Testing recommendations

3. Requesting Changes

Comment on a PR with specific change requests:

/ai-issue-resolver-change Please add error messages below each input field

Configuration

Input	Description	Required	Default
github-token	GitHub API token	Yes	N/A
model-api-key	API key for the Language Model (supports OpenAI, Anthropic, etc.)	Yes	N/A
model-provider	LLM provider choice	No	openai
model-name	Model to use	No	gpt-4
base-branch	Base branch for PRs	No	main
review-threshold	Review confidence threshold	No	0.8

LLM Provider Support Status

Provider	Status	Supported Models	Features
OpenAI	🧪 Testing	gpt-4, gpt-3.5-turbo	All features fully supported
Anthropic (Claude)	🧪 Testing	claude-3.5 sonnet, claude-instant	Basic code generation, review features
Azure OpenAI	🧪 Testing	gpt-4, gpt-3.5-turbo	Basic code generation, review features
Ollama	🚧 Planned	Llama 3.3, codellama, qwen, codegemma	Not yet implemented

Best Practices

1. Use clear, detailed issue descriptions
2. Review AI-generated changes before merging
3. Utilize slash commands for specific updates
4. Provide feedback to improve future generations

Support

Need help? We're here:

1. Check the Documentation
2. Open an Issue
3. Visit our Discussions
4. Contact: support@axiotree.com

Local Testing

To test the action locally:

1. Install Act based on your operating system:

   macOS (using Homebrew):

   brew install act

   Linux (using curl):

   curl -s https://raw.githubusercontent.com/nektos/act/master/install.sh | sudo bash

   Windows (using Chocolatey):

   choco install act-cli

   Alternatively, you can download the latest release from: https://github.com/nektos/act/releases

2. Set up required environment variables:

   export GITHUB_TOKEN=your_github_token
   export MODEL_API_KEY=your_api_key

   For Windows CMD:

   set GITHUB_TOKEN=your_github_token
   set MODEL_API_KEY=your_api_key

   For Windows PowerShell:

   $env:GITHUB_TOKEN = "your_github_token"
   $env:MODEL_API_KEY = "your_api_key"

3. Run the test script:

   chmod +x test/run-local.sh
   ./test/run-local.sh

   For Windows:

   # Using Git Bash or WSL is recommended
   bash test/run-local.sh

This will simulate the GitHub Actions environment locally and test the action with a sample issue event.

Prerequisites

1. Docker Installation

   • Install Docker Desktop (recommended for Windows/Mac)
   • For Linux: sudo apt-get install docker.io
   • Ensure Docker daemon is running before testing

2. Required Tokens

   • GitHub Personal Access Token with repo scope
   • OpenAI / Anthropic / Ollama, etc API Key for AI operations

Troubleshooting

• If you get permission errors, make sure the script is executable: chmod +x test/run-local.sh
• For Windows users, using Git Bash or WSL (Windows Subsystem for Linux) is recommended
• Make sure Docker is installed and running, as Act requires it to simulate the GitHub Actions environment
• If you get "act not found", make sure it's in your system's PATH
• For Docker-related issues:
  • Check if Docker daemon is running: docker ps
  • Start Docker service if needed:
    • Linux: sudo systemctl start docker
    • Windows/Mac: Start Docker Desktop
  • Verify Docker installation: docker --version

License

MIT License - see LICENSE file for details