Composable Orchestration
Overview: Layer 2 — choreograph functions as YAML flows with Event Script.
At a glance
- What — Event Script, a YAML DSL that describes an end-to-end transaction as a flow that choreographs functions — orchestration as configuration, not code.
- Why it matters — changing how a transaction runs means editing a flow, not rewriting Java; roughly half config, half code.
- Where it sits — Layer 2 of the ascent, built on the event-driven foundation.
- For developers and architects orchestrating composable functions.
Event Script is a Domain Specific Language (DSL) that uses YAML to represent an end-to-end transaction flow. A transaction is a business use case, and the flow can be an API service, a batch job or a real-time transaction.
It moves orchestration out of code and into configuration: instead of writing Java that calls one function after another, you describe the sequence as a flow in YAML. Roughly half the work becomes configuration, half stays code (the functions themselves). A flow builds on the event-driven foundation beneath it — it never references a function's class, only its route name.
The mental model
- A flow is an ordered set of tasks with one entry point (
first.task) and one or more terminal tasks. - Each task runs a function (named by route) — a task is a function in its flow-step role. Input and output data mapping move values between the flow and the function's scope.
- A task's execution type decides what happens next:
sequential,decision,parallel,fork,pipeline,response,end, orsink. - A per-transaction state machine (
model) holds intermediate results across the stateless functions. - A flow adapter (HTTP via
rest.yaml, or Kafka) drives a flow from outside and returns its result.
See also
- Event Script Syntax — the full DSL: flow files, writing a flow end to end, every task type, and the data-mapping mini-language.
- Flow grammar — the rule-based schema (the deterministic spec).
- Flow AI agent guide — author flows deterministically with an agent.
- Flow Configuration Schema — exhaustive field reference.
- Event-driven Foundation — the Layer-1 core flows build on (down the ascent).
- Knowledge Graph as Application — the semantic layer above (Layer 3).
- Composing the layers — how the three layers stack.