Runnable Examples

The repository includes small, focused examples for topology, state contracts, runtime control, and multi-agent patterns.

Agent workflows

These examples call a model. Configure an OpenAI-compatible provider with Model Providers before running them.

go run ./examples/graph
go run ./examples/plan_mode -profile multi-step
go run ./examples/supervisor_mode

Workflows with explicit state bindings

go run ./examples/two_agent_handoff "Research explicit state bindings."
go run ./examples/multi_llm_turns "Draft and review an explanation."
go run ./examples/shared_tool_loop "Use the calculator to evaluate 125 * 48."

Runtime control without a model

go run ./examples/graph/conditional_routing.go
go run ./examples/graph/dynamic_map_reduce.go
go run ./examples/graph/failure_fallback.go
go run ./examples/graph/human_approval.go
go run ./examples/graph/fan_in_fan_out.go

Start with the model-free examples when learning checkpoint, failure, approval, and parallel merge behavior. They run quickly and do not require model credentials.

Read the source alongside the run

Goal	Start here
Conditional routing	examples/graph/conditional_routing.go
Dynamic fan-out and reducers	examples/graph/dynamic_map_reduce.go
Failure classification and fallback	examples/graph/failure_fallback.go
Human approval and resume	examples/graph/human_approval.go
Static fan-in/fan-out	examples/graph/fan_in_fan_out.go
Plan/execute/review	examples/plan_mode/
Supervisor/worker orchestration	examples/supervisor_mode/
Graph JSON and registry schema	examples/dsl/

Most model-driven examples read provider settings from the process environment. Keep .local/ output out of source control; remove it when you need a fresh Session or checkpoint history.