Orchestrates the complete lifecycle of Betty agents by managing manifests, system prompts, and skill execution.
The Betty Agent Runner is a sophisticated execution engine designed to power specialized AI agents within the Betty framework. It manages the entire agent workflow, from loading YAML manifests and generating Claude-optimized prompts to orchestrating API calls and executing planned skills. By supporting both iterative reasoning—where agents adapt based on feedback—and oneshot execution for predictable tasks, it provides a robust foundation for building autonomous development tools. The skill includes comprehensive JSON logging, diagnostic error handling, and a built-in mock mode for rapid development without API dependencies.
Key Features
01Full agent lifecycle orchestration from manifest loading to execution
02Support for iterative and oneshot reasoning modes for flexible task handling
030 GitHub stars
04Automated translation of agent manifests into optimized system prompts
05Comprehensive execution logging with structured JSON output for auditability
06Built-in mock simulation mode for testing workflows without Claude API calls
Use Cases
01Executing complex multi-step developer workflows like API design and specification validation
02Automating repetitive coding tasks using specialized agent personas and personas
03Debugging agent reasoning and skill selection using detailed execution logs
