Docs
Documentation
Everything you need to build and run workflows on your own machine — start with the basics, then dive into the node reference.
Getting started
Core concepts
Deterministic-first
Why most steps should be plain and predictable, with AI added on purpose.
ReadAI opt-in per step
How AI steps work, when to use one, and why they stay contained.
ReadLocal brain, cloud spine
Where work runs, where state lives, and what never leaves your machine.
ReadBring your own auth
How Destaris uses your own credentials for AI, email, and providers.
ReadLoops
The third process object — a graph you design that repeats under a small brain until a success condition you wrote is genuinely met.
ReadAgentic Workflows
See what your AI is doing, not another black box — the Agentic Workflow object, its memory, and the Agentic Workflow⇄Workflow lifecycle behind Destaris.
ReadBuilding workflows
The canvas
A tour of the workflow editor — rail, palette, canvas, inspector.
ReadNodes & the palette
The categories of steps you build workflows from.
ReadVisual ↔ YAML
The canvas and the file are two views of one workflow.
ReadDedupe & caching
Only act on what's new — never process the same content twice.
ReadExpressions (JSONata)
How to reference and reshape data with JSONata expressions.
ReadAI authoring skill
The Claude Code skill Destaris installs so AI can author your workflows.
ReadBuilding Loops
How to author a Destaris Loop — the pinned input and brain nodes, the success condition, stage-addressed feedback, and the traps worth knowing before your first run.
ReadBuilding agentic workflows
Loop engineering in Destaris — the trigger, verifier, and stop rules around an AI step, as plain deterministic YAML.
Read