Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Check whether a paper reports each item from PRISMA, CONSORT, STROBE, ARRIVE, SPIRIT, or TRIPOD (per whichever study-design-tool-gate dispatched to), citing where each item is or isn't addressed — including a/b sub-item hierarchy where the standard defines one. Use this after study-design-tool-gate has dispatched to one of these 6 reporting standards; this checks report completeness (did they say where), not methodological quality (was the study done well) — there is no overall synthesis step, j
.claude/skills/yogsoth-ai-reporting-standard-checklist/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | -43% | 0% |
| case-07 | ✓→✓ | ＝ Same ✓ | 83% | 0% |
| case-08 | ✓→✓ | ＝ Same ✓ | 206% | 0% |
| case-09 | ✓→✓ | ＝ Same ✓ | 154% | 0% |
| case-01 | ✗→✗ | ＝ Same ✗ | 329% | 0% |
Per-item report-completeness check (with location citation) across PRISMA/CONSORT/STROBE/ARRIVE/SPIRIT/TRIPOD. No integration step — unlike quality-appraisal-checklist, judgment per item IS the terminal output.
Subagent — spawned via spawn-agent skill.
references/item-sets.md — the authorial-to-reader reversal note that applies to all 6 standards is there, read before drafting (Progressive Disclosure — kept out of this SKILL.md body).
<!-- BEGIN available-tables (generated) -->
| SOP | When to use | | --- | --- | | spawn-agent | Spawn a customized CC subagent with full MCP tool access. |
<!-- END available-tables (generated) -->
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 12,437 | 33,405 | +169% | 1 | 1 | 0% | 1,322 | 5,675 | +329% | 0 | 0 | — |
case-02 | fail→fail | 10,275 | 23,296 | +127% | 1 | 1 | 0% | 888 | 929 | +5% | 0 | 0 | — |
case-03 | fail→fail | 25,797 | 49,170 | +91% | 1 | 1 | 0% | 4,058 | 8,425 | +108% | 0 | 0 | — |
case-04 | fail→fail | 14,920 | 38,278 | +157% | 1 | 1 | 0% | 1,795 | 6,283 | +250% | 0 | 0 | — |
case-05 | fail→fail | 50,281 | 52,810 | +5% | 1 | 1 | 0% | 8,234 | 8,417 | +2% | 0 | 0 | — |
case-06 | fail→fail | 15,078 | 25,887 | +72% | 1 | 1 | 0% | 1,758 | 1,789 | +2% | 0 | 0 | — |
case-07 | pass→pass | 17,723 | 26,291 | +48% | 1 | 1 | 0% | 2,149 | 3,935 | +83% | 0 | 0 | — |
case-08 | pass→pass | 17,487 | 37,990 | +117% | 1 | 1 | 0% | 1,977 | 6,057 | +206% | 0 | 0 | — |
case-09 | pass→pass | 20,636 | 39,380 | +91% | 1 | 1 | 0% | 3,027 | 7,695 | +154% | 0 | 0 | — |
case-10 | fail→pass | 15,296 | 9,481 | -38% | 1 | 1 | 0% | 1,579 | 899 | -43% | 0 | 0 | — |
case-11 | fail→fail | 17,366 | 25,744 | +48% | 1 | 1 | 0% | 2,355 | 819 | -65% | 0 | 0 | — |
case-12 | fail→fail | 50,835 | 33,603 | -34% | 1 | 1 | 0% | 7,702 | 4,946 | -36% | 0 | 0 | — |
case-13 | fail→fail | 26,237 | 59,524 | +127% | 1 | 1 | 0% | 3,427 | 8,404 | +145% | 0 | 0 | — |
case-14 | fail→fail | 16,950 | 23,658 | +40% | 1 | 1 | 0% | 1,993 | 465 | -77% | 0 | 0 | — |
case-15 | fail→fail | 44,842 | 21,854 | -51% | 1 | 1 | 0% | 7,517 | 1,106 | -85% | 0 | 0 | — |
case-16 | fail→fail | 10,652 | 22,683 | +113% | 1 | 1 | 0% | 917 | 1,118 | +22% | 0 | 0 | — |
case-17 | fail→fail | 10,240 | 21,620 | +111% | 1 | 1 | 0% | 821 | 1,477 | +80% | 0 | 0 | — |
case-18 | fail→fail | 26,998 | 26,800 | -1% | 1 | 1 | 0% | 3,542 | 1,463 | -59% | 0 | 0 | — |
case-19 | fail→fail | 10,606 | 19,017 | +79% | 1 | 1 | 0% | 796 | 469 | -41% | 0 | 0 | — |
case-20 | fail→fail | 12,141 | 56,606 | +366% | 1 | 1 | 0% | 1,343 | 6,884 | +413% | 0 | 0 | — |
case-21 | fail→fail | 14,678 | 22,567 | +54% | 1 | 1 | 0% | 1,511 | 2,962 | +96% | 0 | 0 | — |
case-22 | fail→fail | 14,884 | 35,085 | +136% | 1 | 1 | 0% | 1,874 | 6,108 | +226% | 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 16 counted toward the lift figure. The other 6 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 16 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.