A Python tool to check your requirements.txt file for package updates, with optional AI-powered changelog analysis and file caching for better performance.
- Check for available updates in your requirements.txt and pyproject.toml files
- Show update severity (major/minor/patch)
- Display package homepages and changelogs when available
- AI-powered changelog analysis - Analyze upgrade safety with Claude, Gemini, OpenAI, or custom AI providers
- Codebase-aware recommendations - AI scans your code to provide specific, actionable advice
- Optional file caching for faster repeated checks
- Support for comments and inline comments in requirements.txt
- Ignores pre-release versions (alpha, beta, release candidates)
Install from PyPI:
pip install req-update-checkTo use AI-powered analysis, install with AI providers:
# Install with all AI providers (Claude, Gemini, OpenAI)
pip install req-update-check[ai]
# Or install from source
git clone https://github.com/ontherivt/req-update-check.git
cd req-update-check
pip install -e ".[ai]"Check for updates without AI analysis:
req-update-check requirements.txtAnalyze upgrade safety with AI (requires API key):
# Analyze a specific package with Claude (default)
export ANTHROPIC_API_KEY="sk-ant-..."
req-update-check requirements.txt --ai-check requests
# Analyze all outdated packages
req-update-check requirements.txt --ai-check
# Use a different AI provider
export GEMINI_API_KEY="..."
req-update-check requirements.txt --ai-check --ai-provider gemini
# Use OpenAI
export OPENAI_API_KEY="sk-..."
req-update-check requirements.txt --ai-check --ai-provider openaireq-update-check [-h] [--no-cache] [--cache-dir CACHE_DIR]
[--ai-check [PACKAGE]] [--ai-provider {claude,gemini,openai,custom}]
[--ai-model MODEL] [--api-key API_KEY]
requirements_fileArguments:
requirements_file: Path to your requirements.txt or pyproject.toml file
Note: pyproject.toml support requires Python 3.11+
General Options:
--no-cache: Disable file caching--cache-dir CACHE_DIR: Custom cache directory (default:~/.req-check-cache)
AI Analysis Options:
--ai-check [PACKAGE]: Analyze updates with AI (optionally specify package name, or analyze all if omitted). Will only display selected package.--ai-provider {claude,gemini,openai,custom}: Choose AI provider (default: claude)--ai-model MODEL: Override default model for the provider--api-key API_KEY: Provide API key directly (or use environment variables)
Basic output:
File caching enabled
The following packages need to be updated:
requests: 2.28.0 -> 2.31.0 [minor]
Pypi page: https://pypi.python.org/project/requests/
Homepage: https://requests.readthedocs.io
Changelog: https://requests.readthedocs.io/en/latest/community/updates/#release-history
With AI analysis:
File caching enabled
The following packages need to be updated:
requests: 2.28.0 -> 2.32.5 [minor]
Pypi page: https://pypi.python.org/project/requests/
Homepage: https://requests.readthedocs.io
🤖 Analyzing with AI...
AI ANALYSIS:
────────────────────────────────────────────────────────────
✅ Safety: SAFE (Confidence: high)
Model: claude-3-5-sonnet-20241022
Tokens: 8,245 in / 1,823 out / 10,068 total
Recommendations:
1. Review the changelog for security fixes in versions 2.29.0-2.32.0
2. Test SSL certificate verification in your application
3. Update request timeout handling if using default timeouts
New Features:
• Improved connection pooling performance
• Better support for modern TLS versions
• Enhanced cookie handling
Summary: This is a safe minor version upgrade with important security
fixes and performance improvements. No breaking changes detected in your
current usage patterns.
────────────────────────────────────────────────────────────
When you use --ai-check, the tool:
- Fetches changelogs from GitHub releases, direct changelog URLs, or package metadata
- Scans your codebase to find how you're using the package
- Sends to AI with context about your usage patterns
- Returns analysis with:
- Safety assessment (safe/caution/breaking)
- Breaking changes that affect your code
- Deprecations in your current usage
- Actionable upgrade recommendations
- Relevant new features
- Token usage statistics
| Provider | Model | Cost/Analysis* | Setup |
|---|---|---|---|
| Claude (Anthropic) | claude-3-5-sonnet-20241022 | ~$0.05 | export ANTHROPIC_API_KEY="sk-ant-..." |
| Gemini (Google) | gemini-2.0-flash-exp | ~$0.01 | export GEMINI_API_KEY="..." |
| OpenAI | gpt-4o | ~$0.05 | export OPENAI_API_KEY="sk-..." |
| Custom | Your choice | Varies | Configure via config file |
*Estimated cost based on typical changelog and codebase size
Option 1: Environment Variables (Recommended)
# For Claude
export ANTHROPIC_API_KEY="sk-ant-..."
# For Gemini
export GEMINI_API_KEY="..."
# For OpenAI
export OPENAI_API_KEY="sk-..."Option 2: Command Line
req-update-check requirements.txt --ai-check --api-key "your-key-here"Option 3: Config File (Coming in Phase 4)
# ~/.config/req-update-check/config.toml
[ai.api_keys]
claude = "sk-ant-..."
gemini = "..."AI analysis results are cached for 24 hours to save on API costs. The cache is automatically invalidated when:
- Your codebase changes (files using the package are modified)
- 24 hours have passed
- You use
--no-cache
The tool supports file caching to improve performance when checking multiple times. You can configure the cache storage:
req-update-check --cache-dir ~/.your-cache-dir requirements.txtThe tool supports requirements.txt files with the following formats:
package==1.2.3
package == 1.2.3 # with spaces
package==1.2.3 # with inline comments
# Full line comments
Note: Currently only supports exact version specifiers (==). Support for other specifiers (like >=, ~=) is planned for future releases.
You can also use req-update-check as a Python library:
from req_update_check import Requirements
from req_update_check.ai_providers import AIProviderFactory
# Basic usage without AI
req = Requirements('requirements.txt', allow_cache=False)
req.check_packages()
req.report()
# With AI analysis
provider = AIProviderFactory.create(
provider_name='claude',
api_key='sk-ant-...', # or set ANTHROPIC_API_KEY env var
)
req = Requirements(
'requirements.txt',
ai_provider=provider,
)
req.check_packages()
# Analyze specific package
req.report(ai_check_packages=['requests'])
# Or analyze all packages
req.report(ai_check_packages=['*'])To set up for development:
- Clone the repository
- Create a virtual environment:
python -m venv venv - Activate the virtual environment:
source venv/bin/activate(Unix) orvenv\Scripts\activate(Windows) - Install development dependencies:
pip install -e ".[dev,ai]"
# Run all tests
python -m unittest
# Run specific test file
python -m unittest tests.test_req_cheq
# Run with coverage
coverage run -m unittest discover
coverage report
coverage xml# Check code style
ruff check .
# Format code
ruff format .
# Auto-fix issues
ruff check --fix .src/req_update_check/
├── ai_providers/ # AI provider implementations
│ ├── base.py # Abstract base class and AnalysisResult
│ ├── claude.py # Claude (Anthropic) provider
│ ├── gemini.py # Gemini (Google) provider
│ ├── openai.py # OpenAI provider
│ ├── custom.py # Custom/local provider
│ └── factory.py # Provider factory
├── ai_analyzer.py # Main analysis orchestrator
├── changelog_fetcher.py # Fetch changelogs from various sources
├── code_scanner.py # Scan codebase for package usage
├── prompts.py # AI prompt templates
├── formatting.py # Output formatting
├── auth.py # API key management
├── cache.py # File caching
├── core.py # Main Requirements class
├── cli.py # Command-line interface
└── exceptions.py # Custom exceptions
Contributions are welcome! Please feel free to submit a Pull Request.
This project is licensed under the MIT License - see the LICENSE file for details.