init

interview-prep/sessions/README.md

@@ -0,0 +1,43 @@
+# Interview Sessions
+
+Machine-readable interview transcripts, one `.md` file per round. The
+`interview/debrief` and `interview/practice` modes write files here automatically
+after a real or practice round. Downstream analysis modes read them; each consumer
+documents its own usage.
+
+## Format
+
+Speaker labels — `**Interviewer:**` / `**Candidate:**` — so a consumer can read
+either side without re-inferring who spoke:
+
+```markdown
+---
+company: Acme Corp
+role: Instructional Designer
+round: behavioral
+date: 2026-06-01
+interviewer_role: Senior HR Partner
+source: debrief
+---
+
+## Q1
+**Interviewer:** Tell me about a time you...
+<!-- competency: stakeholder-management -->
+**Candidate:** ...answer...
+```
+
+`round`: `screen | hiring-manager | technical | system-design | behavioral | onsite | final`.
+`source`: `debrief | practice | mock | manual`.
+
+## Competency tags (optional)
+
+A `<!-- competency: tag[, tag...] -->` comment on the line directly above a
+`**Candidate:**` line annotates that answer's competency — lowercase-kebab-case,
+comma-separated for multiple. Optional; a consumer that wants tags can infer them
+when absent.
+
+## Privacy — important
+
+Sessions contain real interviewer names and companies. This directory is
+gitignored (only this README and `.gitkeep` are tracked) — session content never
+enters version control.