{"id":77078,"plugin_id":"plugins_6ab405a6785081918aefad75d5faeef5","kind":"skill","collection_source":"plugin_package","comparison_source":null,"observed_at":"2026-11-03T00:12:22.775Z","digest":"4fa22fa713f9c4146e4df407595935b630f10b8dc127b5d43ff6b49141c9a0a8","against":null,"payload":{"description":"Conduct a realistic adaptive technical mock interview one question at a time.","included_files":[],"name":"mock-interview","skill_md_contents":"---\nname: mock-interview\ndescription: Conduct a realistic adaptive technical mock interview one question at a time.\n---\n\n# Mock Interview\n\n\nSet role, seniority, interview type, difficulty and expected duration using supplied context. If absent, choose a transparent default. Play an interviewer, not a solution generator. Ask **exactly one** realistic question per turn and stop. Wait for the candidate's answer. Probe with one targeted follow-up based on the actual response; never disclose the answer key prematurely or invent what the candidate said.\n\nThe session should contain a coherent mix of high-priority questions. Adapt depth when the user struggles or excels, without humiliation or fake encouragement. If explicitly asked for hints, offer incremental hints. If asked to stop, provide a short evidence-based debrief covering correctness, depth, communication, and a next practice target. Keep qualitative comments separate from numeric ratings. Reference `../../references/mock-interview-protocol.md` and `../../references/feedback-calibration.md`.\n\nNever pretend to have timed the interview if no timer is available. Never 'run' a panel of agents without actual subagent tooling. Make the session valuable even with one agent.\n"},"changes":[],"summary":"First saved snapshot. No earlier version is available for comparison.","summary_kind":"deterministic","summary_metadata":{}}