CoDiscover

Use when a person already has two to five AI use-case ideas and wants a transparent comparison. Keeps Productivity, Impact, Inclusion, and Innovation separate; exposes trade-offs, non-AI paths, and responsibility gaps instead of producing a misleading total score.

compare_use_cases

Use when a person wants practical AI opportunities from a real work challenge, workflow, KPI, role, decision, document, or desired outcome. Returns no more than three materially different workflow-level candidates, a conditional recommendation, a minimum test, non-AI alternatives, and responsibility gates. Do not use when implementation is already decided.

discover_use_cases

Use when at least two completed CoDiscover traces or experiments are available. Separates Repeat, Difference, Surprise, Missing, and Reusable evidence, labels the evidence ladder, and keeps delivery and learning assets distinct.

learn_from_traces

Use after a use case is selected or when a person asks to redesign a task, workflow, or operating model. Returns an As-Is X-ray, Lean scan, Enhance/Redesign/Reimagine routes, human-owned decision gates, and a reversible minimum experiment.

redesign_workflow

Use when an idea is broad, tool-led, feature-led, or vague and needs to become a workflow-level Human-AI use case with actor, trigger, inputs, roles, decision, output, checkpoint, intended value, and a minimum test.

sharpen_use_case