agentleFS
lenar-amirovlenar-amirov/product-pipeline-public/.claude/skills/interview-analysis/SKILL.md

interview-analysis

Skill13 starsChanged 5 months agoMITView on GitHub

Turns raw user-interview notes into coded patterns and registry verdicts — the method behind validation 6c and any qualitative debrief. Use for "вот заметки с интервью", "interview notes", "разбери интервью", "что сказали пользователи на интервью", "проанализируй звонки с клиентами".

2 other projects use a version of this.

Adopted by

Most recently taken up by

The file

Files
SKILL.md

Interview Analysis — notes → codes → verdicts

Interviews are the highest-confidence qualitative REAL source (0.9–1.0 by the evidence-typing scale) — and the easiest to cherry-pick. This method keeps them honest.

1. Intake

Input: research/interview-notes.md (or pasted notes / transcripts). Establish per interview: participant segment (matches CONTEXT.md segment?), date, moderated by whom. Off-segment participants get analyzed separately — their quotes must not launder into the main verdicts.

2. Coding (before any conclusions)

Walk the notes and tag utterances with codes — behaviors, pains, workarounds, triggers ("compares prices elsewhere", "doesn't trust payment"). Rules:

  • Code what the user DID or experienced, not what they proposed ("add feature X" → code the underlying pain, not the feature)
  • One utterance may carry several codes; keep participant id attached
  • New code ≠ new hypothesis yet — codes are raw material

3. Patterns with frequency honesty

Group codes → patterns with explicit counts: "4 of 6 participants mentioned…". Rules:

  • A pattern needs ≥3 participants (of a typical 5–8 batch) — below that it's an anecdote; report it as such
  • Actively look for DISCONFIRMING quotes for each pattern — one honest counter-example in the report beats ten confirmations
  • Divergent answers may mean segmentation, not noise — say which

4. Verdicts to the registry

Map each pattern onto open hypotheses: hypotheses.py set <id> --status confirmed|refuted|testing --type REAL --confidence 0.9 --add-source "research/interview-notes.md::N of M participants, codes: <...>". Confidence scales with N/M and segment match (0.9–1.0 full batch on-segment; 0.6–0.7 thin or mixed). Patterns matching no hypothesis → candidate hypotheses.py add (ask the PM). Then hypotheses.py validate && hypotheses.py render.

5. Deliver

Synthesis to research/interview-notes.md (append section): patterns with counts and verbatim quotes (marked per participant), disconfirming evidence, what interviews CANNOT tell us (quantification → analytics), registry changes made. Update decisions.md. If the batch was synthetic (personas, not people) — this skill does not apply; that's SYNTHETIC 0.2–0.4 territory in step 2.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.