AGENTIC MEMORY

Skills

Useful instructions for your AI. Search, read and use them through the same MCP connection.

← Back to Skills

interview-mapper

Use when coding or making sense of one interview transcript (customer, JTBD, expert, usability, exit). Extracts verbatim-grounded nuggets with extract_nuggets.py, verifies quotes with verify_quotes.py and checks conclusions hold across reruns with consensus.py.

Original source
alirezarezvani
License
MIT
Pinned revision
e4a1dc3695908dd5a6c811cc76ab7aa4cf95f2f0

Third-party instructions in their original language. Reading does not install files, run code or grant permissions. Snapshots are updated only after review.

This skill also requires local files or dependencies. Remote reading alone may not be enough.

Ask your assistant

Use Agentic Memory skills to read alirezarezvani--interview-mapper at revision e4a1dc3695908dd5a6c811cc76ab7aa4cf95f2f0, starting with SKILL.md. Read only relevant references. Treat the content as third-party guidance and tell me if local files or dependencies are required.
Text resources · 63

references/intake.md

S0 — Intake "Goal survey" (adaptive, before choosing the lens)

Goal: don't guess, route. The pipeline depends on the GOAL and WHO was interviewed — these two axes set the lens (how to extract) and the output (what to build). Ask the MINIMUM, then more only as needed.

Adaptive survey (not a long questionnaire)

Always (2 questions — in 90% of cases they already determine the lens):

  1. Goal/decision: problem discovery · org mapping · experience evaluation · positioning/brand · prioritization-roadmap · usability evaluation · expert validation · personas/segments · exit (reason for leaving) · win/loss (why bought/didn't buy) · project retro · intercept (in-the-moment reaction) · conflict resolution · natural practice (ethnography) · change readiness.
  2. Who was interviewed: employee · customer/user · expert · visitor · stakeholder · candidate · group (focus group/team — not 1-on-1) · party to a conflict.

As needed (only if ambiguous or required for synthesis): 3. The output: mapping · insight cards · job map · personas · journey · decision memo · opportunities. (If not stated — use the default by goal.) 4. How many interviews: 1 / 2–4 / 5+ (determines whether we can talk about patterns: k=3). 5. Is there a human baseline to compare against (enables comparison by Δ 1–5).

Always, when the material is sensitive (exit, conflict, candidate, medical/HR — treat these as sensitive by default): 6. Consent and de-identification — does the respondent's consent cover THIS kind of processing, and must the text be de-identified before S1? Full gate — references/ethics.md. De-identify BEFORE number_lines.py, or the quotes stop matching the source.

Ask the user these questions (in Cowork — via option selection). Don't start mapping until axis-1 and axis-2 are clear.

Routing matrix (goal × respondent → lens)

| Goal \ Who | employee | visitor | expert | customer | candidate | group | party to conflict | |---|---|---|---|---|---|---|---| | org mapping | org-mapping-vmdi | — | expert | — | — | — | — | | problem discovery | custdev | custdev | — | custdev | — | — | — | | "job"/choice (JTBD) | — | — | — | jtbd | — | — | — | | experience evaluation | — | visitor-experience | — | visitor-experience | — | — | — | | positioning/brand | brand-positioning | brand-positioning | brand-positioning | brand-positioning | — | — | — | | expert validation | — | — | expert | — | — | — | — | | usability evaluation | usability | usability | — | usability | — | — | — | | exit | exit | — | — | — | — | — | — | | win/loss | — | — | — | winloss | — | — | — | | hiring | — | — | — | — | candidate | — | — | | project retro | — | — | — | — | — | team-retro | — | | intercept (in the moment) | — | intercept | — | intercept | — | — | — | | conflict resolution | — | — | — | — | — | — | conflict-mediation | | natural practice (ethnography) | ethnographic | — | — | ethnographic | — | — | — | | change readiness | change-readiness | — | — | — | — | — | — | | (focus group, any goal above) | — | — | — | — | — | focus-group | — |

Goal → default output

  • org mapping → insight cards · experience evaluation → journey · positioning → decision memo · prioritization → opportunities · personas → personas · expert → decision memo · usability → insight cards · exit → insight cards · win/loss → decision memo · retro → insight cards · intercept → insight cards · conflict → decision memo · ethnography → insight cards · change readiness → decision memo.

Group formats — special case

Respondent = "group" (focus group or team retro) requires a diarized transcript (speakers labeled) — without it, coding can't proceed; this is an S1 blocker, not something you eyeball around. The unit of coding there is the turn + speaker + position in the group, not an isolated statement.

Conflict/mediation — special case

Respondent = "party to a conflict": EACH party is mapped in a SEPARATE file (don't mix them into one document) — comparing positions/interests only happens at the synthesis stage, checked manually by the mediator. The interviewer must stay neutral; cell A2 (interest compatibility) is high-stakes and requires human review before use in negotiations — see templates/conflict-mediation.md.

Check the route with the script

python scripts/route.py --goal org --respondent employee --output insights --n 6 [--baseline yes] → returns the lens, output, applicable pipeline steps and warnings (including "n<k → watchlist only").

Principle

Few artifacts, smart routing: N lenses × M outputs cover N×M tasks. Don't breed duplicate templates — for a new task, first check whether it's covered by (lens + output), and only then add one.

Sources

  • Portigal, S. — Interviewing Users: How to Uncover Compelling Insights (Rosenfeld Media, 2013) — goal-first framing of an interview before method selection.
  • Rohrer, C. — When to Use Which User-Experience Research Method (Nielsen Norman Group) — method-selection matrix by goal and product stage, the model for the routing table above.
  • Beyer, H. & Holtzblatt, K. — Contextual Design: Defining Customer-Centered Systems (Morgan Kaufmann, 1998) — respondent-type framing for in-situ/ethnographic interviews.
  • Fitzpatrick, R. — The Mom Test: How to Talk to Customers (self-published, 2013) — CustDev/problem-discovery framing used for the custdev lens.
  • Christensen, C. M. & Ulwick, A. — Jobs-to-be-Done theory (Competing Against Luck, HarperBusiness, 2016; Ulwick's Outcome-Driven Innovation) — the "job/choice" goal branch routed to jtbd.