Build your first multi-model workflow
Choose one outcome, assign bounded roles, and run a coherent staged board.
Task-based documentation
Start with the thing you need to accomplish. Every guide states prerequisites, ordered steps, success evidence, warnings, and the next relevant path.

Start here
Choose one outcome, assign bounded roles, and run a coherent staged board.
Save one encrypted key per provider and reuse it across assigned boxes.
Workflow architecture
Match the orchestration topology to the dependency structure of the work.
Use role controls as behavior, not decoration.
Give the supervisor explicit review, stop, takeover, and team rules.
Load roles, stages, and effort without resetting the user’s provider and model choices.
Local inference
Start the loopback companion and pair it with the same computer’s browser.
Start and pair the device-local bridge without creating a site-owned model endpoint.
Route normal box inference through the paired user-owned model; reserve GoldCap for training work.
GoldCap training
Create a canonical local dataset from explicitly approved pairs or saved-workflow scope.
Launch a user-configured toolkit and track only the process states the companion can observe.
Diagnosis
Match the visible text to a known failure before changing credentials, models, or hardware.
Export and embed
Include or omit DecaCap attribution per export without changing the underlying workflow.
Place an authored recipe diagram on another site without exposing keys, prompts, or private outputs.