structured-elicitation

To extract expert knowledge for an agent configuration, conduct a structured user interview across five layers to capture operating rhythms, recurring decisions, dependencies, institutional knowledge, and friction, then generate agent-ready artifacts for downstream consumption.

Structured elicitation for agent onboarding is a five-layer interview process with checkpoints that translates tacit work patterns into machine-readable configurations, enabling rapid onboarding of new agents or client teams by capturing operating rhythms and institutional knowledge.

You can create an operating model from user interviews by guiding users through a structured five-layer elicitation to verify captured data at each checkpoint, then producing an operating-model.json file that defines recurring decisions and dependencies.

Extracting agent configurations generates multiple machine-readable artifacts including SOUL.md, USER.md, HEARTBEAT.md, operating-model.json, and schedule-recommendations.json to define agent personas, operating rhythms, and recurring decisions.

Yes, you can refine an existing agent configuration by conducting a structured elicitation interview to capture updated operating rhythms, dependencies, and institutional knowledge, then regenerating agent-ready artifacts to align the persona with current workflows.

The best way to document institutional knowledge for downstream agents is to use a five-layer interview with verification checkpoints, ensuring tacit work patterns and dependencies are accurately translated into machine-readable artifacts like USER.md.