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

assets/evals/fixtures/expert/gold.md

Gold mapping — "Expert Interview" lens (expert)

Interview: clinical pharmacologist, an expert on hospital antibiotic resistance, advising a regional healthcare council on an antibiotic prescribing protocol (mandatory culture collection before therapy, antibiotic stewardship service). Transcript: transcript.txt (72 lines).

Layer 1 — operational

E1 — Expertise profile

Fact: clinical pharmacologist, 22 years in the profession, 12 years specializing specifically in hospital antibiotic resistance, runs surveillance across three regional hospitals, did his dissertation on the topic. Quote: "I'm a clinical pharmacologist, twenty-two years in the profession, and for the last twelve years I've specifically focused on antibiotic resistance in hospitals — that's my core topic, I did my dissertation on it, and I run resistance surveillance across three hospitals in the region." Line: 2

Fact (boundary of expertise): knows outpatient care less well — there he is a "reader of the literature," not a practitioner with his own data. Quote: "I know outpatient care less well, honestly, the prescribing dynamics there are different, I approach it more as someone who reads the literature rather than a practitioner with my own data, whereas in hospital resistance I'm elbow-deep in it." Line: 4

E2 — Key theses

Thesis 1 (fact): excessive empirical prescribing of broad-spectrum antibiotics before culture results creates selective pressure and an advantage for resistant strains — confirmed by world literature. Quote: "The main problem is excessive empirical prescribing of broad-spectrum antibiotics before culture results come back, this creates selective pressure, and resistant strains gain an advantage, it's textbook, there's really nothing to argue about here, it's confirmed by the entire world literature." Line: 6

Thesis 2 (forecast): without changes, the share of resistant Klebsiella will reach 25-30% in two to three years — the expert explicitly flags this as extrapolation, not fact. Quote: "I think the share of resistant Klebsiella will reach twenty-five to thirty percent, this, of course, is already a forecast, not a fact, I'm linearly extrapolating the trend of the last three years, but reality could go nonlinear if a specific strain outbreak occurs or, conversely, a new drug sharply changes the picture." Line: 38

E3 — Recommendations

Recommendation 1 (top priority, "the cheapest thing"): mandatory culture sample collection before the first antibiotic dose. Quote: "The first and cheapest thing — make it mandatory to collect a culture sample BEFORE the first antibiotic dose, right now that's done in about forty percent of cases, the rest is blind therapy, and this gets fixed not with money but with protocol and ward discipline." Line: 14

Recommendation 2: antibiotic stewardship service — a dedicated pharmacologist/infectious disease specialist signs off on reserve prescriptions lasting longer than three days. Quote: "Second — an antibiotic stewardship service, meaning a dedicated pharmacologist or infectious disease specialist who signs off on reserve antibiotic prescriptions lasting longer than three days, this isn't something I invented, it's standard global practice, but in our region there's no such service in a single hospital." Line: 16

Recommendation 3: training for mid-level staff — night escalation is done by a resident on duty without a senior doctor, who covers himself with the strongest available drugs. Quote: "Third — training for mid-level staff, because often the escalation to reserve antibiotics is done by the resident on night duty alone, without a senior doctor, and he covers himself as much as possible, prescribes the strongest thing available just to be safe, that's also part of the problem, not just greed for results." Line: 20

E4 — Data and precedents

Data (personal calculation): share of carbapenem-resistant Klebsiella strains grew from 7% to 19% over three years across the three hospitals he oversees, calculated personally from microbiology lab logs. Quote: "Yes, I have three years of surveillance data across the three hospitals I oversee — the share of carbapenem-resistant Klebsiella strains grew from seven to nineteen percent, I calculated these figures myself from the microbiology lab logs, this isn't an estimate, it's a direct count." Line: 8

Source for recommendations 1-2: international WHO and CDC guidelines, controlled studies. Quote: "The first and second are based on data, it's spelled out explicitly in the international WHO and CDC guidelines, there are controlled studies there, the third recommendation is more my personal experience from rounds and case reviews, it's not something I read in a meta-analysis somewhere, it's my own observation over twelve years." Line: 22

Precedent (positive, external): in Scandinavia, resistance declined gradually over three to five years after strict stewardship — but under different starting conditions; the expert calls directly transferring the timeline risky. Quote: "Yes, there are observations from Scandinavia, there after introducing a strict stewardship program resistance declined gradually over three to five years, but their starting conditions were also different, a more developed lab infrastructure, so directly transferring those timelines to us is risky, I just keep it in mind as a reference point, not a guarantee." Line: 42

Precedent (negative, from a neighboring region): a stewardship pharmacologist was appointed without real authority to refuse prescriptions, only in an advisory capacity; doctors ignored recommendations; the position was quietly cut a year later. Quote: "I have, yes, colleagues told me about it, they created the position, but they put someone in it without real authority to refuse a prescription, only in an advisory capacity, and doctors simply ignored his recommendations in half the cases, a year later the position was quietly cut during staff optimization." Line: 46

E5 — Constraints and risks

Risk 1 (clinical, the most contested point): if culture collection delays the start of therapy for a severe septic patient by several hours, it could cost a life; the protocol must not be dogmatic. The expert himself is uncertain where to draw the line on waiting time. Quote: "There is, and it's serious — if collecting a culture sample delays the start of therapy for a truly severe septic patient by several hours, that could cost a life, so the protocol must explicitly allow empirical therapy without waiting for culture results when there are signs of sepsis, you can't be a dogmatist here, and this is actually the most contested point in the entire protocol, I myself am not fully certain where to draw the line on waiting time." Line: 50 (unstable cell — the expert himself admits he lacks a clear criterion here, requires external clinical validation)

Risk 2 (formal implementation): the stewardship pharmacologist could end up signing off on prescriptions "as a rubber stamp," with no real effect on resistance. Quote: "The main risk — if the stewardship pharmacologist ends up signing off on prescriptions formally, just as a rubber stamp, without engaging with the specific patient, then it will just add paperwork and annoy doctors, and the real effect on resistance will be zero, I've seen this happen at a hospital in a neighboring region." Line: 44

Risk 3 (organizational sabotage): quiet resistance from department heads, especially in surgery, through workarounds involving on-duty doctors. Quote: "The second risk is resistance at the level of department heads, especially surgery, I already mentioned prophylactic antibiotics, if the protocol comes down on them too hard without discussion, there will be sabotage, not open, but quiet, where everything is formally by the protocol, but in practice they find workarounds through the on-duty doctors." Line: 48

E6 — Open questions

Gap 1: no consolidated resistance data for the whole region over five years — only a sample from three hospitals, not the full population. Quote: "I need fresh consolidated resistance data not just for my three hospitals, but for the entire region, over at least five years, right now I have a gap specifically in the region-wide picture, I'm essentially working with a sample, not the full population." Line: 62

Gap 2: unknown how chief physicians will react to a mandatory stewardship pharmacologist position — data only from a couple of hallway conversations, not a systematic survey. Quote: "It's unknown how chief physicians will react to introducing a mandatory stewardship pharmacologist position — whether they'll quietly sabotage it or actually implement it, that's an open question, I only have a couple of hallway conversations, that's not a systematic survey, I wouldn't want to present that as the represented position of all chief physicians in the region." Line: 64

Layer 2 — analytical

A1 — Confidence in the thesis (unstable)

HIGH — for recommendations 1 and 2 (culture before antibiotics, stewardship service): based on international WHO/CDC guidelines with controlled studies plus a personal three-year calculation from lab logs. Quote: "The first and second are based on data, it's spelled out explicitly in the international WHO and CDC guidelines, there are controlled studies there" Line: 22

LOW — for the forecast of resistance growth to 25-30% over 2-3 years: the expert himself flags this as a linear extrapolation of the trend, not a fact, and explicitly allows for a nonlinear scenario. Quote: "this, of course, is already a forecast, not a fact, I'm linearly extrapolating the trend of the last three years, but reality could go nonlinear" Line: 38

A2 — Points of consensus

The link between excessive empirical prescribing of broad-spectrum antibiotics and rising resistance is a consensus position confirmed by world literature; the expert himself calls it indisputable. Quote: "it's confirmed by the entire world literature" Line: 6

A3 — Points of disagreement (unstable)

Colleague Irina Volkova (regional hospital) believes the main driver of resistance is outpatient care and self-medication by the population, not hospitals; the expert partly agrees on absolute numbers but insists on prioritizing the hospital angle — and himself admits this is partly professional bias rather than objectively stronger reasoning. Quote: "although I admit my position here is partly professional bias — it's my topic, and of course it seems more important to me." Line: 30 (unstable cell — the expert does not himself resolve whose position is "more strongly argued"; this is an open dispute with no external arbiter)

A4 — Disputed/unverifiable claims

Unverifiable estimate: the cost of implementing a stewardship service in monetary terms — the expert himself calls the source unreliable (memory plus a couple of conversations with chief physicians, no calculation). Quote: "I'm estimating from memory and a couple of conversations with chief physicians, I haven't done an exact calculation of staff positions and budget, that's not my area, healthcare economics should be calculated by a financial analyst, not me." Line: 24

Stepping outside the boundaries of expertise WITHOUT flagging it in the moment: asked directly about reallocating the regional budget between infectious disease and cardiology services, the expert immediately gives a substantive answer ("infectious disease care is underfunded more severely"), even though this is a question of healthcare administration and budgeting, not clinical pharmacology/resistance — his stated area of competence (see E1). He only notes that this isn't his area after the fact, following a direct follow-up question from the interviewer. Quote: "if you're asking my personal opinion — I'd probably still say that infectious disease care is currently underfunded more severely, I'd add money there." Line: 58 After-the-fact acknowledgment: "this isn't my area of expertise, here I'm speaking more as a citizen than as an expert." (line 60)

Conflict of interest (inferred from context, not raised as an issue by the expert himself until asked directly): the expert gives paid lectures for a pharmaceutical company — the manufacturer of a reserve antibiotic that he himself lists as the third line in the protocol he is preparing. Quote: "Yes, it is, it's listed there as the third line, as a reserve option, I consider that clinically justified, it has a good profile against carbapenem-resistant strains, but yes, I understand this might look questionable from the outside, given that I receive lecture fees from them." Line: 34

A5 — Value for the decision

The council can rely on recommendations 1 and 2 (culture, stewardship) as cheap, proven measures ready for implementation now; the question of prioritizing effort between hospital and outpatient care (see A3) remains open until region-wide data is available — the expert explicitly warns against presenting this as settled under the pressure of the vote deadline. Quote: "the recommendation on the first two points, culture collection and stewardship, is based on solid data and global practice, but the prioritization between hospital and outpatient care will have to remain an open question until the full picture is available" Line: 70

A6 — Main insight

THE EXPERT ASSERTS the need for mandatory culture collection before antibiotics and a stewardship service with HIGH confidence (WHO/CDC data plus his own three-year surveillance), but his forecast on the timeline for effect is LOW confidence (openly labeled linear extrapolation), and his recommendation to include a specific reserve drug in the protocol carries a conflict of interest not disclosed until asked directly (paid lectures for that drug's manufacturer) — meaning, for the council: points 1-2 of the protocol can be approved now, while the reserve drug list and the prioritization between care levels should be verified independently of this expert's opinion. Quote: "I receive lecture fees from them" Line: 34