GitHub Issue Summarizer

简体中文 | English

A CLI tool to fetch, filter, and summarize GitHub issues, with one-line summaries generated by LLMs (e.g., OpenAI).

---

🚀 Quick Start (Zero Hassle)

1. Install dependencies:

   pip install -r requirements.txt

2. Run with one command:

   python llm-issue.py owner/repo --max-issues 30

   • Replace owner/repo with the target GitHub repository, e.g. AstrBotDevs/AstrBot.
   • --max-issues controls how many open issues to fetch (default 50, recommended for large repos).
   • No need to manually export environment variables!
   • If GH_TOKEN or OPENAI_API_KEY is missing, you will be prompted to enter them once, and they will be saved to .env for future runs.

3. Results:

   • Output files are saved to the output/ directory automatically.
   • Includes a Markdown summary and filtered issues in JSON.

---

Features

• Fetches open issues from any GitHub repository
• Filters out assigned, closed, or noisy issues
• Classifies issues by type and priority
• Generates one-line summaries for each issue using LLM (e.g., OpenAI GPT)
• Outputs a Markdown table and project overview
• Caches API responses and summaries for better performance
• Supports batch processing with automatic rate limiting
• Quality control for generated summaries
• Fallback to local summarization when LLM is unavailable

Advanced Usage

Optional Arguments

• --max-issues N Limit the number of open issues to fetch (default 50). Useful for large repositories to speed up the process.

Optional Environment Variables

You can still use environment variables or .env to customize behavior:

# Caching options
CACHE_DB_PATH=.cache/cache.db         # Cache database location
CACHE_MAX_MEMORY_ITEMS=1000           # Max items in memory cache
CACHE_CLEANUP_INTERVAL=3600           # Cache cleanup interval (seconds)

# LLM options
LLM_CONCURRENCY_LIMIT=10              # Max concurrent LLM requests
OPENAI_BASE_URL=https://api.openai.com/v1  # API endpoint
MODEL_NAME=gpt-3.5-turbo              # Model to use

Configuration

• Edit config.py to customize:
  • Keywords for issue classification
  • Priority rules and labels
  • LLM prompt templates
  • Filter rules for issues

Output Format

The tool generates:

1. A project overview with issue statistics
2. A detailed Markdown table with:
   • Issue number and type
   • Priority level
   • One-line summary
   • Key labels
   • Creation date
   • Direct link

Example output:

# owner/repo Issues Overview

Currently has **50** open issues (20 bugs / 15 feature requests), average priority P1, last updated on 2024-03-20.

| #Issue | Type | Priority | Title | Summary | Key Labels | Created | Link |
|--------|------|----------|-------|---------|------------|---------|------|
| #123 | Bug | P1 | Login fails | "Login page error under high concurrency" | backend, critical | 2024-03-19 | 🔗 |

Dependencies

• Python 3.8+
• httpx
• typer
• pydantic
• rich
• openai
• python-dotenv

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

Apache License 2.0