Version: 0.8.2
SPARC CLI is a powerful command-line interface that implements the SPARC Framework methodology for AI-assisted software development. It combines autonomous research capabilities with guided implementation to provide a comprehensive toolkit for analyzing codebases, planning changes, and executing development tasks.
- Framework Integration: Seamlessly implements SPARC Framework's methodology
- Autonomous Capabilities: Provides both research and guided implementation
- Safety Controls: Features human-in-the-loop controls and review mechanisms
- Provider Flexibility: Supports multiple LLM providers
- Development Workflow: Enhances productivity through AI assistance
- Research and implementation modes
- Multiple LLM provider support
- Interactive chat mode
- Human-in-the-loop interaction
- Expert knowledge queries
- Shell command execution
- Rich console output
- Web scraping and data extraction
The CLI includes powerful web scraping capabilities:
- Configurable rate limiting and request throttling
- Automatic retry mechanisms
- Response caching
- Error handling and recovery
- Support for multiple data formats (HTML, JSON, XML)
Example usage:
sparc scrape --url "https://example.com" --selector "article.content" --format json- Usage Guide - Installation and basic usage
- Advanced Features - Detailed features and configurations
- Architecture - System design and components
For installation and getting started, please refer to the Usage Guide.
Environment variables are automatically managed in $HOME/.sparc_exports and sourced in your shell's RC file.
Required API Keys:
- ANTHROPIC_API_KEY - Get it from: https://console.anthropic.com/account/keys
Optional API Keys:
- OPENAI_API_KEY - Get it from: https://platform.openai.com/api-keys
- OPENROUTER_KEY - Get it from: https://openrouter.ai/keys
- GEMINI_API_KEY - Get it from: https://makersuite.google.com/app/apikey
- VERTEXAI_PROJECT and VERTEXAI_LOCATION - Configure in Google Cloud Console
The installation script supports both local Python and Docker-based installations. Choose the installation method that best suits your needs:
./install.sh
./install.sh --docker