What is an Entry Point?
The Entry Point defines which node executes first when your pipeline runs—the “starting line” of your workflow.Key Rules
Single Entry Point:- Exactly one node must be designated as the entry point
- Pipeline execution always begins at this node
- ✔️ Can be entry points: LLM, Agent, Toolkit, MCP, Code, Custom, State Modifier, Printer, Decision
- ✘ Cannot be entry points: Router node
Router nodes require input data to evaluate their conditions. At pipeline start, no state variables have been populated yet, so the router has nothing to evaluate.
Setting an Entry Point
Visual Method (Flow Editor)
Steps:- Open your pipeline in the Flow Editor
- Click the three dots (⋮) on the node card
- Select Make entrypoint from the dropdown

YAML Method
In YAML configuration, define the entry point at the top level:Best Practices
Match Entry Point to Workflow Type
- Conversational: Start with LLM node
- Automated: Start with Toolkit or MCP node
- AI-powered routing: Start with Decision node
- Data processing: Start with Code node
- User interaction: Start with Printer node (for displaying initial message)
- State setup: Start with State Modifier node
- Node Connectors - Connect nodes to create workflows
- State Management - Understand state in pipelines
- Flow Editor - Visual pipeline building
- Control Flow Nodes - Router and Decision nodes
- YAML Configuration - Complete YAML syntax reference