What problem does it solve? Service-design personas often become invented character sketches with demographics and interiority that no evidence supports. This Skill produces a persona where every statement traces back to a Discovery insight or an intent field, so design decisions rest on evidence rather than imagination. ## Core Features & Use Cases - Evidence-traced persona generation: Builds identity, goals, behaviors, frustrations, and would_leave_if entries strictly from the intent, Discovery, and ecosystem documents, with frustrations linked to insight ids. - Schema-validated output: Resolves the live persona schema via legion document list, drafts against its required fields, validates with legion document validate, and creates the document with legion document create. - Agent-as-customer support: Handles ecosystems where the primary actor is an agent, writing the agent persona and the unaware human beneficiary separately. - Uncertainty instrumentation: Emits one staked prediction per persona via legion uncertainty emit, with confidence anchors based on trace quality, witnessed later by the crit step. - Use Case: After landing a Discovery and ecosystem for a CLI product, invoke this Skill to produce a draft persona document whose frustrations each cite a supported insight id, ready for crit review. ## Quick Start Ask the agent to write a persona for the primary actor in the landed ecosystem, tracing every statement to the Discovery insights and intent.
