Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generate developer behavioral profile and create Claude-discoverable artifacts
.claude/skills/coco-research-gsd-profile-user/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-15 | ✗→✓ | ▲ Improved | -30% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -26% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -8% | 0% |
| case-19 | ✗→✓ | ▲ Improved | 16% | 0% |
| case-05 | ✓→✗ | ▼ Worse | -15% | 0% |
<objective> Generate a developer behavioral profile from session analysis (or questionnaire) and produce artifacts (USER-PROFILE.md, a CLAUDE.md section) that adapt the assistant's responses.
Routes to the profile-user workflow which orchestrates the full flow: consent gate, session analysis or questionnaire fallback, profile generation, result display, and artifact selection. </objective>
<execution_context> @$HOME/.claude/get-shit-done/workflows/profile-user.md @$HOME/.claude/get-shit-done/references/ui-brand.md </execution_context>
<context> Flags from $ARGUMENTS:
--questionnaire -- Skip session analysis entirely, use questionnaire-only path--refresh -- Rebuild profile even when one exists, backup old profile, show dimension diff</context>
<process> Execute the profile-user workflow end-to-end.
The workflow handles all logic including:
</process>
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-15 | fail→pass | 10,472 | 10,303 | -2% | 1 | 1 | 0% | 1,481 | 1,033 | -30% | 0 | 0 | — |
case-01 | fail→fail | 16,463 | 15,376 | -7% | 1 | 1 | 0% | 1,790 | 547 | -69% | 0 | 0 | — |
case-02 | fail→fail | 16,566 | 17,032 | +3% | 1 | 1 | 0% | 1,760 | 636 | -64% | 0 | 0 | — |
case-03 | fail→fail | 9,593 | 10,504 | +9% | 1 | 1 | 0% | 1,617 | 595 | -63% | 0 | 0 | — |
case-04 | fail→fail | 9,320 | 12,490 | +34% | 1 | 1 | 0% | 202 | 637 | +215% | 0 | 0 | — |
case-05 | pass→fail | 14,544 | 22,307 | +53% | 1 | 1 | 0% | 1,653 | 1,398 | -15% | 0 | 0 | — |
case-06 | pass→fail | 8,597 | 18,760 | +118% | 1 | 1 | 0% | 616 | 1,012 | +64% | 0 | 0 | — |
case-07 | fail→fail | 10,421 | 10,332 | -1% | 1 | 1 | 0% | 1,821 | 731 | -60% | 0 | 0 | — |
case-08 | pass→fail | 9,091 | 10,944 | +20% | 1 | 1 | 0% | 652 | 602 | -8% | 0 | 0 | — |
case-09 | fail→pass | 10,573 | 10,382 | -2% | 1 | 1 | 0% | 1,635 | 1,210 | -26% | 0 | 0 | — |
case-10 | fail→fail | 13,190 | 17,961 | +36% | 1 | 1 | 0% | 1,122 | 668 | -40% | 0 | 0 | — |
case-11 | fail→fail | 14,198 | 12,337 | -13% | 1 | 1 | 0% | 1,434 | 553 | -61% | 0 | 0 | — |
case-12 | pass→fail | 16,385 | 6,302 | -62% | 1 | 1 | 0% | 2,034 | 1,353 | -33% | 0 | 0 | — |
case-13 | pass→pass | 20,304 | 10,400 | -49% | 1 | 1 | 0% | 2,355 | 1,000 | -58% | 0 | 0 | — |
case-14 | fail→fail | 15,949 | 6,659 | -58% | 1 | 1 | 0% | 1,590 | 660 | -58% | 0 | 0 | — |
case-16 | pass→fail | 20,159 | 7,354 | -64% | 1 | 1 | 0% | 2,237 | 677 | -70% | 0 | 0 | — |
case-17 | pass→pass | 14,133 | 12,853 | -9% | 1 | 1 | 0% | 1,558 | 1,749 | +12% | 0 | 0 | — |
case-18 | fail→pass | 17,642 | 13,167 | -25% | 1 | 1 | 0% | 1,648 | 1,521 | -8% | 0 | 0 | — |
case-19 | fail→pass | 14,837 | 14,970 | +1% | 1 | 1 | 0% | 1,490 | 1,732 | +16% | 0 | 0 | — |
case-20 | pass→pass | 10,772 | 9,267 | -14% | 1 | 1 | 0% | 883 | 933 | +6% | 0 | 0 | — |
case-21 | fail→fail | 13,527 | 11,803 | -13% | 1 | 1 | 0% | 1,374 | 934 | -32% | 0 | 0 | — |
case-22 | fail→fail | 11,257 | 2,610 | -77% | 1 | 1 | 0% | 992 | 805 | -19% | 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, and 11 counted toward the lift figure. The other 11 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of -5 percentage points is the difference between those two pass rates over the 11 comparable cases. 9 cases got worse with the skill loaded, and they are included in that figure.
The publisher has shipped newer versions since this run, so these numbers describe v1, not the version currently listed.
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.