BugOut

Automated bug fix workflow that fetches GitHub issues, extracts features, generates PRDs, creates fixes, finds reviewers, and prepares patch folders.

Overview

BugOut performs 8 automated steps:

Fetch Comments - Uses gh CLI to get all issue comments and saves as JSON
Feature Extraction - Uses AI (parser.py style) to extract structured features from comments
PRD Generation - Analyzes features and generates a Product Requirements Document
Bug Fix Generation - Uses AI to propose a fix based on the PRD
Reviewer Check - Uses Yutori API to find competent reviewers from issue commenters
Patch Folder - Prepares a complete patch folder with all artifacts
Repo Clone & Agentic Loop - Clones the repo and uses OpenAI to analyze and generate precise code changes
Patch Creation - Generates actual unified diff patch files and updates the directory

Requirements

Python 3.8+
GitHub CLI (gh)
.env file in project root with:
- FASTINO_KEY - For AI inference
- YUTORI_KEY - For reviewer competence checking
- OPENAI_HOST - OpenAI API host (e.g., api.openai.com)
- OPENAI_MODEL - OpenAI model to use (e.g., gpt-4o)
- OPENAI_API_KEY - OpenAI API key (optional for some endpoints)
wxPython - For GUI (optional, see GUI section)

Installation

CLI Installation

# Install dependencies
pip install requests python-dotenv

# Ensure gh CLI is installed
gh --version

# Create .env file in project root (parent of qwen/)
cat > ../.env << EOF
FASTINO_KEY=your-key
YUTORI_KEY=your-key
OPENAI_HOST=api.openai.com
OPENAI_MODEL=gpt-4o
OPENAI_API_KEY=your-openai-key
EOF

GUI Installation (Optional)

# Install all dependencies including wxPython
pip install -r requirements.txt

# Or install wxPython separately:

# Linux (Ubuntu/Debian)
sudo apt-get install python3-wxgtk4.0

# macOS
pip install wxPython

# Windows
pip install wxPython

Usage

Full Workflow (CLI)

Run the complete BugOut workflow from command line:

cd qwen
python bugout.py <repo> <issue_number> [output_dir]

Examples:

# Basic usage
python bugout.py microsoft/vscode 12345

# With custom output directory
python bugout.py facebook/react 67890 ./my_output

# Show help
python bugout.py --help

Graphical Interface (GUI)

Launch the BugOut graphical interface:

python bugout.py --gui

GUI Features:

Component	Description
📝 Config Panel	Input fields for repo, issue number, output directory
📊 Status Panel	Progress bar, step counter (0/8), Run ID display
📋 Log Panel	Color-coded live output (success/error/warning/info)
🖱️ Control Buttons	Run, Stop, Clear Log
📁 Directory Browser	Browse button for selecting output location
🎯 Visual Indicators	Step-by-step progress with Unicode symbols

GUI Workflow:

Enter repository (e.g., microsoft/vscode)
Enter issue number (e.g., 12345)
Optionally select output directory
Click "🚀 Run BugOut"
Watch real-time progress in the log panel
View completion summary with patch folder location

GUI Screenshots:

╔══════════════════════════════════════════════════════════════╗
║     🐛 BugOut - Automated Bug Fix Workflow              ║
╠══════════════════════════════════════════════════════════════╣
║  From bug report to production-ready patch                    ║
╚══════════════════════════════════════════════════════════════╝

Repository:    [microsoft/vscode________________]
Issue Number:  [12345___________________________]
Output Dir:    [./bugout_data___________________] [Browse...]

Status: Running
Progress: [████████████████░░░░░░░░] 62%  Step: 5/8
Run ID: a3f5b2c1

Output Log:
  ✓ Step 1 complete: Fetched 15 comments for issue #12345
  ✓ Step 2 complete: Extracted features from 16 entries
  ✓ Step 3 complete: Generated PRD with 16 reports analyzed
  ✓ Step 4 complete: Generated bug fix
  ✓ Step 5 complete: Checked 8 reviewers
  → Best reviewer: @somecontributor
  ...

[🚀 Run BugOut]  [⏹ Stop]  [🗑 Clear Log]

Individual Steps

You can also run each step individually:

Step 1: Fetch Comments

python comment_fetcher.py <repo> <issue_number> [output_dir]

Step 2: Extract Features

python feature_extractor.py <comments.json> [output_file]

Step 3: Generate PRD

python prd_generator.py <bugs_with_features.json> [output_prd.md]

Step 4: Generate Bug Fix

python bug_fixer.py <prd.md> <bugs_with_features.json> [output_dir]

Step 5: Check Reviewers

python reviewer_checker_wrapper.py <comments.json> <repo> [output_dir] [wait]

Step 6: Prepare Patch Folder

python patch_generator.py <output_dir> <prd.md> <bug_fix.patch> <reviewer.json> <comments.json> <bugs_with_features.json>

Step 7: Clone Repo & Agentic Loop

python repo_cloner.py <repo> <prd.md> <bug_fix.json|bug_fix.patch> [output_dir]

Step 8: Create Patch File

python patch_creator.py <clone_path> <agent_response.json> <output_dir>

Output Structure

Each run gets a unique 8-character UUID as its directory name for easy tracking and organization.

bugout_data/
└── a3f5b2c1/                 # ← Unique run ID (UUID)
    ├── run_metadata.json     # Run info: UUID, repo, issue, timestamp
    ├── issue_comments.json   # Raw issue data from GitHub
    ├── bugs_with_features.json  # Extracted features
    ├── prd.md                # Product Requirements Document
    ├── prd.analysis.json     # Feature analysis JSON
    ├── bug_fix.patch         # Initial proposed fix
    ├── bug_fix.json          # Fix details JSON
    ├── reviewer.json         # Reviewer analysis
    ├── agent_response.json   # Agentic loop output (Step 7)
    ├── generated.patch       # AI-generated unified diff (Step 8)
    ├── git.patch             # Git diff patch (Step 8)
    ├── applied_changes.json  # Applied changes log (Step 8)
    ├── temp/                 # Temp directory with repo clone
    │   └── microsoft_vscode_clone/  # Cloned repository
    ├── repo_snapshot/        # Snapshot of modified repo
    └── patch/                # Complete patch folder
        ├── prd.md
        ├── bug_fix.patch
        ├── reviewer.json
        ├── issue_comments.json
        ├── bugs_with_features.json
        ├── analysis.json
        ├── bug_fix.json
        ├── agent_response.json
        ├── generated.patch
        ├── git.patch
        ├── applied_changes.json
        └── patch_manifest.json  # Includes run_id reference

Run Metadata

Each run creates a run_metadata.json file:

{
  "run_id": "a3f5b2c1",
  "repo": "microsoft/vscode",
  "issue_number": "12345",
  "timestamp": "2026-02-27T10:30:00.000000",
  "output_dir": "./bugout_data/a3f5b2c1"
}

Generated Artifacts

prd.md

Product Requirements Document containing:

Executive summary
Frequency analysis of bug characteristics
Technical details from user reports
Functional and non-functional requirements
Success criteria

bug_fix.patch

Proposed fix containing:

Root cause analysis
Fix description
Code changes
Testing instructions

reviewer.json

Reviewer analysis containing:

List of all commenters
Competence assessment for each
Best reviewer recommendation

agent_response.json (Step 7)

Output from the agentic loop containing:

Root cause analysis
Affected files list
Fix strategy
Detailed code changes with line numbers
Testing recommendations
Confidence score

generated.patch (Step 8)

Unified diff patch file containing:

All code changes from the agentic loop
Proper diff format for easy review
Can be applied with git apply or patch

git.patch (Step 8)

Git-formatted patch from the actual repository changes:

Generated from git diff HEAD
Ready for git am or PR creation

applied_changes.json (Step 8)

Log of all applied changes containing:

List of changes with success/failure status
Analysis summary from the agent
Testing instructions
Confidence score

Environment Variables

Variable	Description	Required
`FASTINO_KEY`	API key for AI inference	Yes
`YUTORI_KEY`	API key for Yutori reviewer check	Yes
`OPENAI_HOST`	OpenAI API host (e.g., `api.openai.com`)	Yes
`OPENAI_MODEL`	OpenAI model to use (e.g., `gpt-4o`)	Yes
`OPENAI_API_KEY`	OpenAI API key	Optional

The .env file should be in the parent directory (project root).

Example Output

     l         __                       __
  .   .       / /  __ _____ ____  __ __/ /_
   \ /       / _ \/ // / _ `/ _ \/ // / __/
 `/ ! \`    /_.__/\_,_/\_, /\___/\_,_/\__/
 | o:o |              /___/
~| o:o |~
/ \_:_/ \

╔══════════════════════════════════════════════════════════════╗
║     🐛 BugOut - Automated Bug Fix Workflow              ║
╠══════════════════════════════════════════════════════════════╣
║  From bug report to production-ready patch                    ║
╚══════════════════════════════════════════════════════════════╝

Configuration 
  Run ID:      a3f5b2c1
  Repository:  microsoft/vscode
  Issue:       #12345
  Output:      ./bugout_data/a3f5b2c1
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

● [1/8] 🔗 Fetching issue comments
  ✅ Step 1 complete: Fetched 15 comments for issue #12345

● [2/8] ⚙️  Extracting features from comments
  ⚙️ Processing 16 text entries...
  ✅ Step 2 complete: Extracted features from 16 entries

● [3/8] 🎯 Generating PRD
  ✅ Step 3 complete: Generated PRD with 16 reports analyzed

● [4/8] 🐛 Generating bug fix
  ✅ Step 4 complete: Generated bug fix

● [5/8] ★ Checking reviewer competence
  ✅ Step 5 complete: Checked 8 reviewers
  ✅ Best reviewer: @somecontributor

● [6/8] 📁 Preparing initial patch folder
  ✅ Created: ./bugout_data/a3f5b2c1/patch

● [7/8] 🚀 Running agentic loop with OpenAI
  🔀 Cloning microsoft/vscode...
  ✅ Clone: ./bugout_data/a3f5b2c1/temp/microsoft_vscode_clone
  → Agent response: ./bugout_data/a3f5b2c1/agent_response.json

● [8/8] ✨ Generating actual patch file
  Applying 3 changes...
  ✅ Generated: ./bugout_data/a3f5b2c1/generated.patch
  ✅ Updated: ./bugout_data/a3f5b2c1/patch

══════════════════════════════════════════════════════════════
  🚀 BugOut Complete! 
══════════════════════════════════════════════════════════════

╔══════════════════════════════════════════════════════════════╗
║              BugOut Summary                                ║
╠══════════════════════════════════════════════════════════════╣
║  Run ID:     a3f5b2c1                                      ║
║  Repository: microsoft/vscode                              ║
║  Issue:      #12345                                        ║
║  Patch Folder: ./bugout_data/a3f5b2c1/patch                ║
║  Best Reviewer: @somecontributor                           ║
╠══════════════════════════════════════════════════════════════╣
║  Generated Artifacts:                                      ║
║    • prd.md                  (Product Requirements Doc)    ║
║    • bug_fix.patch           (Initial Proposed Fix)        ║
║    • generated.patch         (AI-Generated Patch)          ║
║    • git.patch               (Git Diff Patch)              ║
║    • reviewer.json           (Reviewer Analysis)           ║
║    • agent_response.json     (Agentic Loop Output)         ║
║    • applied_changes.json    (Applied Changes Log)         ║
╠══════════════════════════════════════════════════════════════╣
║  Next Steps:                                               ║
║    1. Review PRD:        cat ./bugout_data/a3f5b2c1/prd.md ║
║    2. Review patch:      cat ./bugout_data/a3f5b2c1/patch/ ║
║                          generated.patch                   ║
║    3. Contact reviewer:  @somecontributor                  ║
║    4. Create PR with the generated patch                   ║
╚══════════════════════════════════════════════════════════════╝

Project Structure

qwen/
├── bugout.py                 # Main CLI orchestrator (8 steps)
├── bugout_gui.py             # wxPython graphical interface
├── comment_fetcher.py        # Step 1: Fetch issue comments
├── feature_extractor.py      # Step 2: AI feature extraction
├── prd_generator.py          # Step 3: Generate PRD
├── bug_fixer.py              # Step 4: Initial bug fix
├── reviewer_checker_wrapper.py # Step 5: Yutori reviewer check
├── patch_generator.py        # Step 6: Initial patch folder
├── repo_cloner.py            # Step 7: Clone repo + agentic loop
├── patch_creator.py          # Step 8: Generate unified diff
├── requirements.txt          # Python dependencies
└── README.md                 # This file

License

MIT

Name		Name	Last commit message	Last commit date
Latest commit History 18 Commits
qwen		qwen
.gitignore		.gitignore
README.md		README.md
bugout.py		bugout.py
comment1.sh		comment1.sh
logo.ansiart		logo.ansiart
parser.py		parser.py
py.py		py.py
requirements.txt		requirements.txt
review_checker.py		review_checker.py
reviewer_check.py		reviewer_check.py

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Repository files navigation

BugOut

Overview

Requirements

Installation

CLI Installation

GUI Installation (Optional)

Usage

Full Workflow (CLI)

Graphical Interface (GUI)

Individual Steps

Output Structure

Run Metadata

Generated Artifacts

prd.md

bug_fix.patch

reviewer.json

agent_response.json (Step 7)

generated.patch (Step 8)

git.patch (Step 8)

applied_changes.json (Step 8)

Environment Variables

Example Output

Project Structure

License

About

Uh oh!

Releases

Packages

Uh oh!

Contributors

Uh oh!

Languages

Folders and files

Latest commit

History

Repository files navigation

BugOut

Overview

Requirements

Installation

CLI Installation

GUI Installation (Optional)

Usage

Full Workflow (CLI)

Graphical Interface (GUI)

Individual Steps

Output Structure

Run Metadata

Generated Artifacts

prd.md

bug_fix.patch

reviewer.json

agent_response.json (Step 7)

generated.patch (Step 8)

git.patch (Step 8)

applied_changes.json (Step 8)

Environment Variables

Example Output

Project Structure

License

About

Resources

Uh oh!

Stars

Watchers

Forks

Releases

Packages 0

Uh oh!

Contributors

Uh oh!

Languages

Packages