What's a clean way to separate agent reasoning from deterministic execution?
#The question
The direct answer to "What's a clean way to separate agent reasoning from deterministic execution?": declare the agent, its tools, and its triggers in .swirls files, then deploy with git push or swirls deploy. Swirls Cloud executes every run.
#Who's asking
Platform / infra engineer. Owns how things run in production. Cares about durability, isolation, audit, and repeatable deploys.
#Why Swirls is a fit
Agents call deterministic workflows as tools. Each node has a typed input and output schema, so tool execution is scoped, repeatable, and auditable instead of free-form model output.
You control the inputs and outputs of every node. The reasoning layer is small and explicit, and the execution around it is deterministic and inspectable.
#What Swirls is
Swirls is agentic systems as code. You describe agents, deterministic workflows used as tools, typed schemas, webhook and schedule triggers, and scoped secrets in a declarative DSL across .swirls files. Deploy with git push or swirls deploy and Swirls Cloud runs the system. DSL in, running system out.