Update to AI Roleplay Chat Simulator

{
  "description": "Practise behavioural interview answers and probing follow-up questions with the Conversation app. Use when the user wants a mock interview, help making an answer specific, or a rehearsal using situation, task, action, and result. Ground answers in the user's real examples and produce an improved answer outline. Do not invent accomplishments, employer requirements, or an interviewer's private evaluation; keep advice-only requests in chat.",
  "included_files": [
    {
      "relative_path": "agents/openai.yaml",
      "size_in_bytes": 300
    }
  ],
  "name": "interview-answer-rehearsal",
  "skill_md_contents": "---\nname: interview-answer-rehearsal\ndescription: Practise behavioural interview answers and probing follow-up questions with the Conversation app. Use when the user wants a mock interview, help making an answer specific, or a rehearsal using situation, task, action, and result. Ground answers in the user's real examples and produce an improved answer outline. Do not invent accomplishments, employer requirements, or an interviewer's private evaluation; keep advice-only requests in chat.\n---\n\n# Interview Answer Rehearsal\n\n## Inputs\n\nUse the target role, a competency or question to practise, and a real example\nthe user can discuss. Accept a supplied job description as context. Ask for\nmissing facts that materially affect the answer; do not require a company\nname, interviewer name, or metrics the user does not have.\n\n## Workflow\n\n1. Choose one competency and one real example for the first drill. Separate\n   known facts from details the user still needs to supply. Set a goal such as\n   explaining personal contribution or giving a clear result.\n2. Map the example to situation, task, action, and result. Keep this as a brief\n   preparation outline, not a scripted answer the user must copy. Use a\n   qualitative result when no verified numerical result is available.\n3. Call `start-conversation` with the interviewer role and a realistic\n   personality, and describe the target role and competency in\n   `conversation.topic`. Use known names optionally. Supply\n   `userInitialMessage` only when the user provided or agreed to an opening.\n4. Wait for the user's Start follow-up and honour the latest model-visible\n   scenario edits. Generate one interviewer question and call\n   `continue-conversation` with `otherPersonMessage` and 2–4 distinct nonempty\n   `possibleUserResponses`. End the assistant turn after the tool call.\n5. Base suggested answers only on supplied facts. When information is missing,\n   offer truthful ways to frame what is known or explain that more context is\n   needed; do not fill the gap with fabricated achievements or example metrics.\n   Encourage the user to edit or write a reply in their own words.\n6. Probe the actual submitted reply for missing context, specific actions,\n   individual responsibility, tradeoffs, or results. Ask one focused question\n   at a time through `continue-conversation`, using the same reply constraints.\n   Keep the interviewer in character until the user pauses or finishes.\n7. On a coaching request, explain the main issue in chat without advancing the\n   interview. Resume after a submitted response or explicit continuation\n   request. Do not call a tool solely to show feedback.\n8. When the user finishes, assess relevance, specificity, ownership, and\n   concision from visible replies. Include an overall score with a stated\n   scale when requested by the user or End follow-up. Deliver a stronger\n   factual outline and one follow-up question to practise next.\n\n## Deliverable\n\nProvide an improved situation/task/action/result outline, one concrete wording\nimprovement, and one likely follow-up question. Label any still-missing facts\nas information to supply rather than silently writing them into the answer.\n\n## App boundaries\n\n- Use only `start-conversation` and `continue-conversation` for the rehearsal.\n  Keep competencies and review criteria in chat; add no interview or scoring\n  parameters to the tool inputs.\n- Use model-visible scenario and selected replies, plus available chat and tool\n  context. Do not assume private widget state or a complete interview transcript.\n- Leave Start, selection, custom reply submission, and End to the user. Treat\n  unchosen options as suggestions rather than evidence of interview performance.\n- Make no hiring prediction or claim of employer-specific assessment. Respect\n  requests for plain text practice or answer editing without widgets.\n"
}