First-Run Welcome

Adaptive onboarding handles cold starts, partial setups, and warm starts by reading knowledge state files like .knowledge/setup-state.yaml to determine the current configuration level and route users to the appropriate guided setup prompts.

You can start a guided onboarding interview by typing /setup to begin the adaptive flow, which reads your active dataset state and routes you through contextual setup prompts before analysis.

User routing during onboarding reads .knowledge/setup-state.yaml, .knowledge/active.yaml, and related manifest or index data to assess the current configuration and decide the appropriate welcome message and guided prompts.

Contextual onboarding supports quick handoff to /setup or question routing depending on user intent, referencing the available dataset actions and current state to guide the session start effectively.

Adaptive onboarding applies during session start and before analysis, ensuring the guided welcome adapts based on the setup state to effectively handle cold starts, partial setups, and warm starts.

The best way to automate guided onboarding is using an adaptive flow that reads setup-state.yaml and active.yaml manifest data to dynamically route users through contextual prompts based on their specific configuration state.