← ClaudeAtlas

persona-interrogatelisted

Interrogate saved personas about a subject - cast each as an isolated agent with its own packet, collect findings and verdicts in a fixed vocabulary, confront disagreements, and synthesise the result. Use to judge whether a plan, document, diff, or decision holds up for the people it affects, or to ask those people questions in character; not for creating personas.
wizspec/agent-workflows · ★ 3 · AI & Automation · score 79
Install: claude install-skill wizspec/agent-workflows
# Persona Interrogate Every run is an interrogation. It takes a subject, a goal stated as questions, and a cast of personas; each persona answers in character from its own perspective, and when the goal is judgement the final question asks for a verdict in a fixed vocabulary. Grading is an interrogation whose last question has a fixed answer set. Read [references/packet.md](references/packet.md) in full before spawning anything. Keep this skill and its reference in the orchestrator; a persona receives only its packet. ## Establish the run Pin down three inputs before choosing anyone: - **the subject**: the plan, document, diff, decision, or other content, as paths or pasted content, plus its stage: a plan, a draft, or shipped work. A plan must not be judged like a shipped document, so the stage goes into every packet. Infer it when the subject makes it obvious and ask when it does not; - **the goal**: what the user wants to know, turned into questions in the order they should be answered. When the user wants a judgement, add "Would you accept this as it is?" as the last question. When they only want answers, leave the verdict out; and - **the cast**: which personas take part. Ask the user only when a missing input would change the run materially. ## Choose the cast When the user names personas, cast those. Otherwise read `docs/wizspec/personas/README.md` and match the subject and goal against each persona's interests; cast the ones that match. Cast everyone