011 GitHub stars
02Dual-Mode Architecture (Agent & Router) for flexible chat ownership
03Resilience-first Execution with bounded retries, exponential backoff, failover, and circuit breakers
04Unified Browser Automation and Intelligence Tools for evidence capture, failure analysis, and workflow verification
05Advanced LLM Tooling including consensus queries, prompt templates, structured output, and response caching
06Multi-Provider Support for OpenAI, GLM, Chutes, Anthropic, Azure OpenAI, and custom OpenAI-compatible endpoints