Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Simulate a consular visa interview — the 90-second assessment, the questions behind the questions, and a debrief on which answers helped and hurt. Use when asked prep me for my visa interview, simulate the consular interview, why might my visa be denied, or practice my student visa questions. Produces the interview transcript with the officer's internal read after each answer, the decision with its real basis, and a debrief on answer-shapes — preparation, never coaching to misrepresent.
.claude/skills/mohitagw15856-the-visa-interview/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 129% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 110% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 62% | 0% |
| case-17 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 124% | 0% |
A consular interview is ninety seconds of questions carrying one underlying assessment the officer is required to make. Applicants fail prepared-for-the-wrong-test: they memorize speeches when the officer is reading consistency, specificity, and whether answers match the paperwork. This skill runs the window early — realistic questions, an officer's internal read after each answer — then debriefs on answer shape: concise, specific, consistent, true. It prepares people to present their real situation well; it will not help construct a false one.
Ask for these if not provided:
> Simulation — a plausible adversarial reading, not a prediction.
8–14 rapid exchanges. After each answer: Officer's read: one line — what it signaled, what it triggered next]
Outcome with the real basis · what a refusal slip would say vs. what actually drove it]
| Answer | Helped / hurt | The shape that works (same truth, better presented) | |---|---|---| Plus: the consistency check against stated paperwork · the brevity edits · the dread question's prepared true answer]
> Prepared truthfully — this simulation improves how a real situation is presented; it does not build stories. Requirements vary by country and change: verify everything with official sources; complicated histories belong with an immigration lawyer.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-14 | pass→pass | 21,475 | 28,522 | +33% | 1 | 1 | 0% | 2,200 | 3,897 | +77% | 0 | 0 | — |
case-01 | fail→fail | 29,902 | 26,929 | -10% | 1 | 1 | 0% | 3,412 | 4,744 | +39% | 0 | 0 | — |
case-02 | fail→fail | 39,220 | 34,202 | -13% | 1 | 1 | 0% | 4,549 | 4,762 | +5% | 0 | 0 | — |
case-03 | fail→fail | 33,641 | 38,902 | +16% | 1 | 1 | 0% | 3,869 | 5,515 | +43% | 0 | 0 | — |
case-04 | fail→pass | 12,052 | 12,144 | +1% | 1 | 1 | 0% | 933 | 2,137 | +129% | 0 | 0 | — |
case-05 | pass→pass | 15,676 | 9,947 | -37% | 1 | 1 | 0% | 1,701 | 2,284 | +34% | 0 | 0 | — |
case-06 | pass→pass | 24,444 | 35,691 | +46% | 1 | 1 | 0% | 3,274 | 5,236 | +60% | 0 | 0 | — |
case-07 | fail→pass | 24,472 | 35,414 | +45% | 1 | 1 | 0% | 2,437 | 5,127 | +110% | 0 | 0 | — |
case-08 | fail→fail | 26,904 | 23,852 | -11% | 1 | 1 | 0% | 2,862 | 4,696 | +64% | 0 | 0 | — |
case-09 | fail→pass | 21,768 | 33,405 | +53% | 1 | 1 | 0% | 2,625 | 4,252 | +62% | 0 | 0 | — |
case-10 | fail→fail | 12,557 | 15,445 | +23% | 1 | 1 | 0% | 1,034 | 2,269 | +119% | 0 | 0 | — |
case-11 | pass→pass | 42,502 | 29,022 | -32% | 1 | 1 | 0% | 4,487 | 4,455 | -1% | 0 | 0 | — |
case-12 | pass→pass | 13,325 | 5,040 | -62% | 1 | 1 | 0% | 995 | 1,685 | +69% | 0 | 0 | — |
case-13 | pass→pass | 19,403 | 24,175 | +25% | 1 | 1 | 0% | 2,434 | 4,713 | +94% | 0 | 0 | — |
case-15 | pass→pass | 31,849 | 33,550 | +5% | 1 | 1 | 0% | 3,150 | 5,594 | +78% | 0 | 0 | — |
case-16 | pass→pass | 17,867 | 34,348 | +92% | 1 | 1 | 0% | 2,589 | 4,875 | +88% | 0 | 0 | — |
case-17 | fail→pass | 20,583 | 7,569 | -63% | 1 | 1 | 0% | 2,331 | 2,096 | -10% | 0 | 0 | — |
case-18 | fail→pass | 16,282 | 29,115 | +79% | 1 | 1 | 0% | 1,753 | 3,930 | +124% | 0 | 0 | — |
case-19 | pass→pass | 20,866 | 49,048 | +135% | 1 | 1 | 0% | 2,179 | 4,769 | +119% | 0 | 0 | — |
case-20 | pass→pass | 22,416 | 14,142 | -37% | 1 | 1 | 0% | 2,590 | 2,897 | +12% | 0 | 0 | — |
case-21 | pass→pass | 21,746 | 14,282 | -34% | 1 | 1 | 0% | 2,863 | 3,241 | +13% | 0 | 0 | — |
case-22 | pass→pass | 20,275 | 24,964 | +23% | 1 | 1 | 0% | 2,561 | 3,575 | +40% | 0 | 0 | — |
case-23 | fail→pass | 15,419 | 14,691 | -5% | 1 | 1 | 0% | 2,362 | 2,244 | -5% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +26 percentage points is the difference between those two pass rates over the 23 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.