Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Deprecated: forwards to bmad-build. Do not use unless invoked by name
.claude/skills/bmad-code-org-bmad-quick-dev/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | 279% | 0% |
| case-05 | ✗→✓ | ▲ Improved | 153% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -13% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-09 | ✗→✓ | ▲ Improved | -19% | 0% |
{project-root}/_bmad/custom/bmad-quick-dev.toml{project-root}/_bmad/custom/bmad-quick-dev.user.tomlbmad-quick-dev is deprecated. Redirecting to bmad-build. Please use bmad-build in the future., invoke bmad-build exactly once with the user's original input verbatim, then execute no further steps in this shim.{project-root}/_bmad/custom/bmad-quick-dev.toml becomes {project-root}/_bmad/custom/bmad-build.toml.{project-root}/_bmad/custom/bmad-quick-dev.user.toml becomes {project-root}/_bmad/custom/bmad-build.user.toml.bmad-quick-dev is deprecated. Redirecting to bmad-build. Please use bmad-build in the future., invoke bmad-build exactly once with the user's original input verbatim, then execute no further steps in this shim.| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 8,731 | 19,162 | +119% | 1 | 1 | 0% | 1,039 | 3,934 | +279% | 0 | 0 | — |
case-02 | fail→fail | 4,992 | 9,649 | +93% | 1 | 1 | 0% | 379 | 1,883 | +397% | 0 | 0 | — |
case-03 | fail→fail | 10,412 | 5,576 | -46% | 1 | 1 | 0% | 1,407 | 1,032 | -27% | 0 | 0 | — |
case-04 | fail→fail | 4,930 | 44,208 | +797% | 1 | 1 | 0% | 461 | 930 | +102% | 0 | 0 | — |
case-05 | fail→pass | 19,827 | 10,922 | -45% | 1 | 1 | 0% | 1,011 | 2,562 | +153% | 0 | 0 | — |
case-06 | pass→pass | 6,225 | 293,241 | +4611% | 1 | 1 | 0% | 878 | 2,824 | +222% | 0 | 0 | — |
case-07 | fail→pass | 7,938 | 4,598 | -42% | 1 | 1 | 0% | 1,096 | 955 | -13% | 0 | 0 | — |
case-08 | fail→pass | 14,708 | 3,571 | -76% | 1 | 1 | 0% | 1,133 | 1,024 | -10% | 0 | 0 | — |
case-09 | fail→pass | 14,339 | 4,817 | -66% | 1 | 1 | 0% | 1,405 | 1,140 | -19% | 0 | 0 | — |
case-10 | pass→fail | 10,688 | 10,357 | -3% | 1 | 1 | 0% | 1,681 | 2,203 | +31% | 0 | 0 | — |
case-11 | fail→pass | 25,058 | 5,082 | -80% | 1 | 1 | 0% | 1,641 | 1,206 | -27% | 0 | 0 | — |
case-12 | fail→pass | 18,695 | 4,845 | -74% | 1 | 1 | 0% | 1,383 | 1,314 | -5% | 0 | 0 | — |
case-13 | fail→pass | 9,949 | 9,964 | +0% | 1 | 1 | 0% | 1,536 | 1,190 | -23% | 0 | 0 | — |
case-14 | pass→pass | 11,221 | 6,487 | -42% | 1 | 1 | 0% | 1,670 | 1,421 | -15% | 0 | 0 | — |
case-15 | fail→pass | 17,630 | 8,322 | -53% | 1 | 1 | 0% | 2,426 | 1,917 | -21% | 0 | 0 | — |
case-16 | fail→pass | 11,191 | 12,325 | +10% | 1 | 1 | 0% | 1,745 | 1,195 | -32% | 0 | 0 | — |
case-17 | fail→fail | 17,250 | 4,538 | -74% | 1 | 1 | 0% | 1,121 | 1,153 | +3% | 0 | 0 | — |
case-18 | fail→pass | 18,889 | 4,535 | -76% | 1 | 1 | 0% | 1,221 | 897 | -27% | 0 | 0 | — |
case-19 | pass→pass | 8,072 | 2,698 | -67% | 1 | 1 | 0% | 1,077 | 778 | -28% | 0 | 0 | — |
case-20 | fail→pass | 12,654 | 9,260 | -27% | 1 | 1 | 0% | 2,164 | 2,095 | -3% | 0 | 0 | — |
case-21 | fail→pass | 12,299 | 4,443 | -64% | 1 | 1 | 0% | 1,742 | 1,183 | -32% | 0 | 0 | — |
case-22 | pass→pass | 11,776 | 4,992 | -58% | 1 | 1 | 0% | 1,903 | 1,151 | -40% | 0 | 0 | — |
case-23 | pass→pass | 25,505 | 7,894 | -69% | 1 | 1 | 0% | 2,538 | 985 | -61% | 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 +52 percentage points is the difference between those two pass rates over the 23 comparable cases. 1 case got worse with the skill loaded, and it is 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.