Features
Type-safe flow composition with TypeScript
Sequence agent for linear multi-step workflows
Parallel agent for concurrent AI tasks
For each agent for iterative processing
Evaluator agent for grading or validating LLM output
Best of all agent for selecting best result from multiple candidates
One-of agent for branching workflows
Agent helper wrapping Vercel AI SDK generateText
Custom agent support via simple function interface
Nested flows for complex orchestration
Context-based execution with shared agents
Lightweight – minimal abstraction over Vercel AI SDK
Multi-agent orchestration with handoffs
Sandbox Agents for containerized long-running tasks
Agent-as-tool delegation (v0.15.0+)
Realtime Agents with gpt-realtime-2 voice support (v0.17.6+)
Human-in-the-loop mechanisms
Automatic session history management
Built-in tracing for debugging
Provider-agnostic LLM support (100+ models via LiteLLM)
Redis session support (optional, v0.17.6+)
Instructions, tools, and guardrails configuration
Jupyter notebook compatibility
Supports OpenAI Responses and Chat Completions APIs
