Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Summarize the current git repository: branch, recent commits, changed files, and overall health. Use when asked for a repo overview or status check.
.claude/skills/tinyhumansai-git-summary/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | 85% | 0% |
| case-14 | ✗→✓ | ▲ Improved | -10% | 0% |
| case-18 | ✗→✓ | ▲ Improved | -70% | 0% |
| case-08 | ✓→✗ | ▼ Worse | -63% | 0% |
| case-12 | ✓→✗ | ▼ Worse | 87% | 0% |
Produce a concise summary of the current git repository state.
git rev-parse --abbrev-ref HEAD to get the current branch.git log --oneline -10 to list the 10 most recent commits.git status --short to show uncommitted changes.git stash list to check for stashed work.git remote -v to list configured remotes.Present the summary as a structured report:
## Repository Summary
**Branch**: <current branch>
**Remotes**: <list of remotes>
**Recent commits** (last 10):
- <commit hash> <message>
...
**Working tree**: <clean | N files changed>
**Stashes**: <none | N stashes>| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 6,815 | 4,418 | -35% | 1 | 1 | 0% | 1,473 | 468 | -68% | 0 | 0 | — |
case-02 | fail→fail | 8,094 | 4,342 | -46% | 1 | 1 | 0% | 825 | 529 | -36% | 0 | 0 | — |
case-03 | fail→fail | 3,248 | 4,587 | +41% | 1 | 1 | 0% | 428 | 470 | +10% | 0 | 0 | — |
case-04 | fail→fail | 6,932 | 4,784 | -31% | 1 | 1 | 0% | 1,501 | 1,252 | -17% | 0 | 0 | — |
case-05 | fail→fail | 2,968 | 6,150 | +107% | 1 | 1 | 0% | 508 | 571 | +12% | 0 | 0 | — |
case-06 | fail→fail | 4,377 | 9,811 | +124% | 1 | 1 | 0% | 765 | 1,978 | +159% | 0 | 0 | — |
case-07 | pass→pass | 5,956 | 3,989 | -33% | 1 | 1 | 0% | 1,107 | 881 | -20% | 0 | 0 | — |
case-08 | pass→fail | 7,980 | 5,446 | -32% | 1 | 1 | 0% | 1,526 | 569 | -63% | 0 | 0 | — |
case-09 | fail→fail | 9,488 | 3,602 | -62% | 1 | 1 | 0% | 1,986 | 462 | -77% | 0 | 0 | — |
case-10 | fail→pass | 6,695 | 12,417 | +85% | 1 | 1 | 0% | 1,331 | 2,469 | +85% | 0 | 0 | — |
case-11 | fail→fail | 2,231 | 7,369 | +230% | 1 | 1 | 0% | 415 | 922 | +122% | 0 | 0 | — |
case-12 | pass→fail | 6,806 | 11,939 | +75% | 1 | 1 | 0% | 1,284 | 2,403 | +87% | 0 | 0 | — |
case-13 | fail→fail | 5,955 | 4,183 | -30% | 1 | 1 | 0% | 1,012 | 490 | -52% | 0 | 0 | — |
case-14 | fail→pass | 9,464 | 7,565 | -20% | 1 | 1 | 0% | 1,936 | 1,744 | -10% | 0 | 0 | — |
case-15 | pass→fail | 3,883 | 5,155 | +33% | 1 | 1 | 0% | 724 | 619 | -15% | 0 | 0 | — |
case-16 | pass→pass | 6,140 | 5,361 | -13% | 1 | 1 | 0% | 1,100 | 1,256 | +14% | 0 | 0 | — |
case-17 | pass→fail | 9,522 | 3,522 | -63% | 1 | 1 | 0% | 1,984 | 427 | -78% | 0 | 0 | — |
case-18 | fail→pass | 8,015 | 1,540 | -81% | 1 | 1 | 0% | 1,594 | 483 | -70% | 0 | 0 | — |
case-19 | pass→fail | 7,200 | 5,308 | -26% | 1 | 1 | 0% | 596 | 526 | -12% | 0 | 0 | — |
case-20 | fail→fail | 4,025 | 3,750 | -7% | 1 | 1 | 0% | 665 | 535 | -20% | 0 | 0 | — |
case-21 | fail→fail | 2,822 | 4,627 | +64% | 1 | 1 | 0% | 551 | 533 | -3% | 0 | 0 | — |
case-22 | pass→pass | 6,475 | 2,533 | -61% | 1 | 1 | 0% | 1,113 | 672 | -40% | 0 | 0 | — |
case-23 | pass→pass | 6,876 | 6,356 | -8% | 1 | 1 | 0% | 1,366 | 1,431 | +5% | 0 | 0 | — |
case-24 | pass→pass | 2,483 | 2,411 | -3% | 1 | 1 | 0% | 400 | 552 | +38% | 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. 24 cases were attempted, and 12 counted toward the lift figure. The other 12 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 -8 percentage points is the difference between those two pass rates over the 12 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.