Score Interview Responses and Generate Assessment

Assesses interview responses using the same structured scoring logic as resume evaluation, ensuring interviews stay consistent with your hiring criteria. The interview-assessor agent reads filled-in INTERVIEW_NOTES.md or interview transcriptions, scores each response 1-5, checks for ticket-jockey patterns (for managers), and produces an INTERVIEW_ASSESSMENT.md with scored Q&A and a recommendation.

1. Conduct interview using INTERVIEW_NOTES.md template and fill in candidate responses. 2. Run /assess-interviews (or say 'Assess interviews' in Copilot). 3. The interview-assessor reads notes/transcription, applies role-specific scoring rubric (IC: 1-5 per question, Manager: higher standard checking for leadership scope). 4. Generates INTERVIEW_ASSESSMENT.md with scores, exact quotes, average, and recommendation. 5. candidate-summary-distiller creates HR_SUMMARY_INTERVIEW.md. Output: scored interview with recommendation and interview HR summary.

The /assess-interviews command scans evaluation folders for INTERVIEW_NOTES.md, interview notes, or transcription files (.vtt, .srt, markdown). interview-assessor parses Q&A pairs with exact quote extraction, applies role-specific scoring thresholds from config, and checks manager-specific patterns (ticket jockey detection) before producing assessment.