Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Resources for conversational interviews to create learner profiles. Used by project-interviewer agent during /init.
.claude/skills/aiskillstore-project-interview/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-03 | ✗→✓ | ▲ Improved | 29% | 0% |
| case-10 | ✗→✓ | ▲ Improved | -66% | 0% |
| case-11 | ✗→✓ | ▲ Improved | -62% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -48% | 0% |
| case-13 | ✗→✓ | ▲ Improved | -46% | 0% |
This skill provides resources for conducting natural, conversational interviews with learners.
One goal: Gather enough information to create a persona.md that fits the learner.
What to understand:
| Resource | Purpose | | -------- | ------- | | conversation-flow.md | Example conversations for reference | | interview-data-template.md | Template for saving conversation records |
Instead of following rigid rules, act like a good counselor:
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 11,460 | 7,961 | -31% | 1 | 1 | 0% | 1,009 | 689 | -32% | 0 | 0 | — |
case-02 | fail→fail | 4,727 | 7,699 | +63% | 1 | 1 | 0% | 724 | 648 | -10% | 0 | 0 | — |
case-03 | fail→pass | 8,726 | 3,117 | -64% | 1 | 1 | 0% | 498 | 643 | +29% | 0 | 0 | — |
case-04 | pass→pass | 12,824 | 12,820 | -0% | 1 | 1 | 0% | 1,435 | 1,514 | +6% | 0 | 0 | — |
case-05 | pass→pass | 14,294 | 10,065 | -30% | 1 | 1 | 0% | 1,358 | 995 | -27% | 0 | 0 | — |
case-06 | fail→fail | 24,439 | 16,683 | -32% | 1 | 1 | 0% | 3,218 | 2,231 | -31% | 0 | 0 | — |
case-07 | fail→fail | 30,306 | 17,771 | -41% | 1 | 1 | 0% | 5,175 | 2,368 | -54% | 0 | 0 | — |
case-08 | fail→fail | 8,138 | 8,309 | +2% | 1 | 1 | 0% | 1,253 | 1,544 | +23% | 0 | 0 | — |
case-09 | pass→pass | 3,748 | 12,804 | +242% | 1 | 1 | 0% | 585 | 1,599 | +173% | 0 | 0 | — |
case-10 | fail→pass | 41,070 | 11,572 | -72% | 1 | 1 | 0% | 6,547 | 2,248 | -66% | 0 | 0 | — |
case-11 | fail→pass | 14,114 | 7,265 | -49% | 1 | 1 | 0% | 1,412 | 538 | -62% | 0 | 0 | — |
case-12 | fail→pass | 18,158 | 4,400 | -76% | 1 | 1 | 0% | 1,907 | 1,001 | -48% | 0 | 0 | — |
case-13 | fail→pass | 7,712 | 7,198 | -7% | 1 | 1 | 0% | 1,106 | 602 | -46% | 0 | 0 | — |
case-14 | fail→pass | 17,456 | 5,644 | -68% | 1 | 1 | 0% | 1,882 | 1,101 | -41% | 0 | 0 | — |
case-15 | pass→pass | 11,954 | 11,810 | -1% | 1 | 1 | 0% | 1,675 | 1,241 | -26% | 0 | 0 | — |
case-16 | pass→pass | 12,952 | 7,091 | -45% | 1 | 1 | 0% | 1,771 | 1,285 | -27% | 0 | 0 | — |
case-17 | fail→pass | 12,038 | 7,151 | -41% | 1 | 1 | 0% | 965 | 555 | -42% | 0 | 0 | — |
case-18 | fail→pass | 18,251 | 7,786 | -57% | 1 | 1 | 0% | 2,050 | 675 | -67% | 0 | 0 | — |
case-19 | fail→pass | 11,233 | 1,181 | -89% | 1 | 1 | 0% | 807 | 453 | -44% | 0 | 0 | — |
case-20 | fail→fail | 23,167 | 11,139 | -52% | 1 | 1 | 0% | 4,173 | 1,037 | -75% | 0 | 0 | — |
case-21 | fail→fail | 7,726 | 5,048 | -35% | 1 | 1 | 0% | 410 | 1,087 | +165% | 0 | 0 | — |
case-22 | fail→fail | 46,896 | 6,994 | -85% | 1 | 1 | 0% | 8,214 | 1,204 | -85% | 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. 22 cases were attempted. The headline lift of +41 percentage points is the difference between those two pass rates over the 22 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.