Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Turns a decision the user can't fully answer into a questionnaire for someone else to fill in. Use when the user says "turn this into a questionnaire" or needs knowledge pulled out of an expert.
.claude/skills/fradser-to-questionnaire/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | -21% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 76% | 0% |
| case-06 | ✗→✓ | ▲ Improved | 54% | 0% |
| case-13 | ✓→✗ | ▼ Worse | -52% | 0% |
| case-10 | ✓→✗ | ▼ Worse | -52% | 0% |
Turn something the user can't answer alone into a questionnaire — a Markdown document they hand to one person to fill in async, or fill out together over a meeting. The recipient holds knowledge the user lacks; the questionnaire pulls it out of them.
Grill the send, not the subject. Interview the user only about the _send_, which they can always answer: who it goes to, and what they need back. The questions in the document then target the gap between what the recipient knows and what the user needs.
to-questionnaire-<slug>.md in the current directory (slug from the topic) and report the path. Done when the file exists and every item the user named in step 2 is covered by a question.Frame the document as a discovery questionnaire: the user lacks context, the recipient holds it. Order questions most-important-first — async means you may only get one pass — and group them under ## headings by theme once there are more than a handful. Write it using the template below.
<questionnaire-template>
Purpose: why this questionnaire exists and the decision riding on it.
From: <the user> — To: <the recipient> — How your answers will be used: <where they go>
One paragraph orienting a recipient who wasn't in the user's head. Enough to answer well, not a page.
Deadline and rough effort. Partial answers and "I don't know" are useful — flag anything you're unsure of rather than skipping it.
One ## section per theme. Under each, its questions, most-important-first. Every question is one idea — never compound — with an answer stub directly beneath, and a one-line _why this matters_ only where the question could be misread or invite a throwaway answer.
<question-example>
_Why this matters: it decides whether we provision for burst traffic now or defer it._
> </question-example>
A closing catch-all: anything we didn't ask that we should know?
</questionnaire-template>
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-14 | pass→pass | 5,053 | 4,621 | -9% | 1 | 1 | 0% | 854 | 1,390 | +63% | 0 | 0 | — |
case-01 | fail→fail | 15,365 | 6,504 | -58% | 1 | 1 | 0% | 2,501 | 1,582 | -37% | 0 | 0 | — |
case-02 | fail→fail | 18,820 | 6,022 | -68% | 1 | 1 | 0% | 3,197 | 1,709 | -47% | 0 | 0 | — |
case-03 | fail→fail | 17,146 | 5,206 | -70% | 1 | 1 | 0% | 2,600 | 1,512 | -42% | 0 | 0 | — |
case-13 | pass→fail | 15,357 | 6,123 | -60% | 1 | 1 | 0% | 2,942 | 1,417 | -52% | 0 | 0 | — |
case-04 | fail→pass | 12,757 | 5,019 | -61% | 1 | 1 | 0% | 1,929 | 1,519 | -21% | 0 | 0 | — |
case-05 | fail→pass | 8,966 | 12,990 | +45% | 1 | 1 | 0% | 1,519 | 2,669 | +76% | 0 | 0 | — |
case-06 | fail→pass | 12,074 | 11,780 | -2% | 1 | 1 | 0% | 1,766 | 2,721 | +54% | 0 | 0 | — |
case-07 | fail→fail | 14,772 | 7,801 | -47% | 1 | 1 | 0% | 2,390 | 1,708 | -29% | 0 | 0 | — |
case-08 | fail→fail | 15,907 | 5,106 | -68% | 1 | 1 | 0% | 2,358 | 1,495 | -37% | 0 | 0 | — |
case-09 | fail→fail | 13,050 | 4,510 | -65% | 1 | 1 | 0% | 2,340 | 1,330 | -43% | 0 | 0 | — |
case-10 | pass→fail | 18,120 | 4,533 | -75% | 1 | 1 | 0% | 3,035 | 1,471 | -52% | 0 | 0 | — |
case-11 | fail→fail | 13,367 | 5,047 | -62% | 1 | 1 | 0% | 2,344 | 1,251 | -47% | 0 | 0 | — |
case-12 | fail→fail | 18,307 | 4,646 | -75% | 1 | 1 | 0% | 3,145 | 1,546 | -51% | 0 | 0 | — |
case-15 | fail→fail | 1,493 | 10,521 | +605% | 1 | 1 | 0% | 237 | 2,683 | +1032% | 0 | 0 | — |
case-16 | pass→fail | 7,210 | 10,723 | +49% | 1 | 1 | 0% | 1,408 | 2,342 | +66% | 0 | 0 | — |
case-17 | pass→fail | 7,202 | 6,944 | -4% | 1 | 1 | 0% | 1,270 | 1,768 | +39% | 0 | 0 | — |
case-18 | fail→fail | 17,128 | 3,915 | -77% | 1 | 1 | 0% | 2,560 | 1,328 | -48% | 0 | 0 | — |
case-19 | fail→fail | 15,841 | 3,631 | -77% | 1 | 1 | 0% | 2,657 | 1,259 | -53% | 0 | 0 | — |
case-20 | fail→fail | 15,077 | 3,371 | -78% | 1 | 1 | 0% | 2,735 | 1,196 | -56% | 0 | 0 | — |
case-21 | fail→fail | 16,692 | 3,788 | -77% | 1 | 1 | 0% | 2,906 | 1,209 | -58% | 0 | 0 | — |
case-22 | fail→fail | 14,665 | 9,330 | -36% | 1 | 1 | 0% | 2,376 | 2,389 | +1% | 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 -5 percentage points is the difference between those two pass rates over the 22 comparable cases. 6 cases got worse with the skill loaded, and they are included in that figure.
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.