interview-synthesislisted

Incremental customer interview synthesis. Code one to three transcripts at a time into a running evidence base, then regenerate a synthesis report — an overall summary first, then one section per topic area the study cares about (existing flow, pain points, messaging testing, pricing sensitivity, and so on). Tracks saturation so you know when to stop interviewing. Use when the user mentions interview synthesis, customer or user interviews, discovery interviews, coding transcripts, research synthesis, "what did we learn from the interviews", "add this interview", or wants a readout from interview notes. Sized for studies up to roughly 15 interviews.
AgLyx3/interview-synthesis · ★ 1 · AI & Automation · score 65

Install: claude install-skill AgLyx3/interview-synthesis

# Interview Synthesis Turn customer interviews into a synthesis report that stays current as you keep interviewing. Most synthesis workflows assume you finish all your interviews, then analyze. Real discovery doesn't work that way — you talk to two people, learn something, talk to two more. This skill is built for that: evidence accumulates permanently, interpretation is rebuilt from scratch every time you ask for a report. ## The one rule that matters **Append the evidence. Regenerate the interpretation.** Coded excerpts are append-only — once P03's words are on disk with a participant ID, they never change. That includes their code labels: when a code is later split or merged, the retirement is declared in `codebook.md` and `scripts/tally.py` folds the counts at read time, rather than anyone going back to relabel old excerpts. But themes, findings and the report are re-derived from the *entire* evidence base every time, never edited in place. This exists to defeat anchoring. If each new interview only ever gets filed under themes that already exist, the study freezes around whatever the first two people happened to say, and interview 9 that contradicts interview 2 gets quietly absorbed instead of breaking the theme open. Re-deriving is cheap at this scale and keeps the report an honest reading of everything you have. ## Workspace All state lives in one directory, by default `research/<study-slug>/` under the current project. Ask where it should go if it isn't obviou