Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Stage, commit, and push git changes with conventional commit messages. Use when user wants to commit and push changes, mentions pushing to remote, or asks to save and push their work. Also activates when user says "push changes", "commit and push", "push this", "push to github", or similar git workflow requests.
.claude/skills/davila7-git-pushing/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -35% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -30% | 0% |
| case-08 | ✗→✓ | ▲ Improved | -67% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -3% | 0% |
| case-15 | ✗→✓ | ▲ Improved | -39% | 0% |
Stage all changes, create a conventional commit, and push to the remote branch.
Automatically activate when the user:
ALWAYS use the script - do NOT use manual git commands:
bashbash skills/git-pushing/scripts/smart_commit.sh
With custom message:
bashbash skills/git-pushing/scripts/smart_commit.sh "feat: add feature"
Script handles: staging, conventional commit message, Claude footer, push with -u flag.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-17 | fail→fail | 3,450 | 3,566 | +3% | 1 | 1 | 0% | 519 | 372 | -28% | 0 | 0 | — |
case-01 | fail→pass | 3,859 | 1,522 | -61% | 1 | 1 | 0% | 651 | 425 | -35% | 0 | 0 | — |
case-02 | fail→pass | 3,207 | 1,370 | -57% | 1 | 1 | 0% | 538 | 378 | -30% | 0 | 0 | — |
case-03 | fail→fail | 4,732 | 2,359 | -50% | 1 | 1 | 0% | 734 | 464 | -37% | 0 | 0 | — |
case-04 | pass→pass | 8,371 | 5,507 | -34% | 1 | 1 | 0% | 1,733 | 1,245 | -28% | 0 | 0 | — |
case-05 | pass→pass | 6,659 | 5,600 | -16% | 1 | 1 | 0% | 1,408 | 1,327 | -6% | 0 | 0 | — |
case-06 | pass→pass | 6,945 | 3,943 | -43% | 1 | 1 | 0% | 1,272 | 996 | -22% | 0 | 0 | — |
case-07 | fail→fail | 2,385 | 1,989 | -17% | 1 | 1 | 0% | 493 | 402 | -18% | 0 | 0 | — |
case-08 | fail→pass | 6,157 | 1,405 | -77% | 1 | 1 | 0% | 1,165 | 385 | -67% | 0 | 0 | — |
case-09 | fail→fail | 2,671 | 4,200 | +57% | 1 | 1 | 0% | 393 | 421 | +7% | 0 | 0 | — |
case-10 | fail→fail | 4,274 | 4,366 | +2% | 1 | 1 | 0% | 793 | 393 | -50% | 0 | 0 | — |
case-11 | fail→fail | 3,186 | 2,278 | -28% | 1 | 1 | 0% | 531 | 418 | -21% | 0 | 0 | — |
case-12 | fail→fail | 2,666 | 3,652 | +37% | 1 | 1 | 0% | 466 | 394 | -15% | 0 | 0 | — |
case-13 | fail→fail | 1,839 | 3,656 | +99% | 1 | 1 | 0% | 358 | 371 | +4% | 0 | 0 | — |
case-14 | fail→pass | 2,458 | 3,435 | +40% | 1 | 1 | 0% | 411 | 399 | -3% | 0 | 0 | — |
case-15 | fail→pass | 3,583 | 3,455 | -4% | 1 | 1 | 0% | 596 | 365 | -39% | 0 | 0 | — |
case-16 | fail→pass | 4,408 | 2,201 | -50% | 1 | 1 | 0% | 829 | 399 | -52% | 0 | 0 | — |
case-18 | fail→pass | 4,250 | 3,691 | -13% | 1 | 1 | 0% | 656 | 421 | -36% | 0 | 0 | — |
case-19 | fail→fail | 2,310 | 4,571 | +98% | 1 | 1 | 0% | 293 | 364 | +24% | 0 | 0 | — |
case-20 | fail→pass | 4,634 | 1,180 | -75% | 1 | 1 | 0% | 718 | 352 | -51% | 0 | 0 | — |
case-21 | fail→pass | 2,497 | 3,785 | +52% | 1 | 1 | 0% | 219 | 369 | +68% | 0 | 0 | — |
case-22 | fail→fail | 1,828 | 3,939 | +115% | 1 | 1 | 0% | 279 | 355 | +27% | 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 15 counted toward the lift figure. The other 7 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 +41 percentage points is the difference between those two pass rates over the 15 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.