Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Execute generates conventional commit messages using AI. It analyzes code changes and suggests a commit message adhering to the conventional commits specification. Use this skill when you need help writing clear, standardized commit messages, especially a... Use when managing version control. Trigger with phrases like 'commit', 'branch', or 'git'.
.claude/skills/jeremylongshore-generating-conventional-commits/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✓→✗ | ▼ Worse | 43% | 0% |
| case-13 | ✓→✗ | ▼ Worse | 16% | 0% |
| case-20 | ✓→✗ | ▼ Worse | 36% | 0% |
| case-21 | ✓→✗ | ▼ Worse | -25% | 0% |
| case-02 | ✓→✓ | ＝ Same ✓ | 53% | 0% |
Generate conventional commit messages by analyzing staged Git changes, selecting the correct type prefix (feat/fix/refactor/etc.), and producing concise, standards-compliant messages.
Create well-formatted, informative commit messages that follow the conventional commits standard, improving collaboration and automation in your Git workflow. It saves you time and ensures consistency across your project.
feat: add new feature, fix: correct bug).This skill activates when you need to:
User request: "Generate a commit message for these changes."
The skill will:
feat: Implement user authentication.User request: "Create a commit for the bug fix."
The skill will:
fix: Resolve issue with incorrect password reset.This skill integrates with your Git workflow, providing a convenient way to generate commit messages directly within Claude Code. It complements other Git-related skills in the DevOps Automation Pack, such as /branch-create and /pr-create.
The skill produces structured output relevant to the task.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→fail | 3,931 | 7,823 | +99% | 1 | 1 | 0% | 590 | 843 | +43% | 0 | 0 | — |
case-02 | pass→pass | 4,886 | 4,285 | -12% | 1 | 1 | 0% | 853 | 1,307 | +53% | 0 | 0 | — |
case-03 | pass→pass | 11,043 | 10,349 | -6% | 1 | 1 | 0% | 1,141 | 1,512 | +33% | 0 | 0 | — |
case-04 | pass→pass | 7,969 | 5,828 | -27% | 1 | 1 | 0% | 1,245 | 1,623 | +30% | 0 | 0 | — |
case-05 | pass→pass | 8,306 | 8,864 | +7% | 1 | 1 | 0% | 965 | 1,165 | +21% | 0 | 0 | — |
case-06 | pass→pass | 10,210 | 4,114 | -60% | 1 | 1 | 0% | 930 | 1,292 | +39% | 0 | 0 | — |
case-07 | pass→pass | 9,901 | 9,054 | -9% | 1 | 1 | 0% | 900 | 1,224 | +36% | 0 | 0 | — |
case-08 | fail→fail | 13,604 | 6,403 | -53% | 1 | 1 | 0% | 1,456 | 1,663 | +14% | 0 | 0 | — |
case-09 | pass→pass | 11,164 | 4,140 | -63% | 1 | 1 | 0% | 943 | 1,195 | +27% | 0 | 0 | — |
case-10 | pass→pass | 9,063 | 11,367 | +25% | 1 | 1 | 0% | 1,541 | 1,710 | +11% | 0 | 0 | — |
case-11 | pass→pass | 8,951 | 8,865 | -1% | 1 | 1 | 0% | 733 | 1,266 | +73% | 0 | 0 | — |
case-12 | pass→pass | 4,205 | 3,918 | -7% | 1 | 1 | 0% | 705 | 1,099 | +56% | 0 | 0 | — |
case-13 | pass→fail | 9,291 | 15,208 | +64% | 1 | 1 | 0% | 727 | 840 | +16% | 0 | 0 | — |
case-14 | pass→pass | 8,631 | 9,123 | +6% | 1 | 1 | 0% | 597 | 1,269 | +113% | 0 | 0 | — |
case-15 | pass→pass | 10,533 | 10,566 | +0% | 1 | 1 | 0% | 1,015 | 1,536 | +51% | 0 | 0 | — |
case-16 | pass→pass | 9,011 | 3,975 | -56% | 1 | 1 | 0% | 776 | 1,296 | +67% | 0 | 0 | — |
case-17 | fail→fail | 14,038 | 8,427 | -40% | 1 | 1 | 0% | 1,546 | 1,751 | +13% | 0 | 0 | — |
case-18 | pass→pass | 7,403 | 6,469 | -13% | 1 | 1 | 0% | 1,486 | 1,638 | +10% | 0 | 0 | — |
case-19 | pass→pass | 5,013 | 4,150 | -17% | 1 | 1 | 0% | 983 | 1,240 | +26% | 0 | 0 | — |
case-20 | pass→fail | 8,399 | 6,998 | -17% | 1 | 1 | 0% | 617 | 839 | +36% | 0 | 0 | — |
case-21 | pass→fail | 9,933 | 2,229 | -78% | 1 | 1 | 0% | 1,045 | 779 | -25% | 0 | 0 | — |
case-22 | pass→pass | 12,524 | 6,281 | -50% | 1 | 1 | 0% | 1,712 | 1,904 | +11% | 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 21 counted toward the lift figure. The other 1 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 -100 percentage points is the difference between those two pass rates over the 21 comparable cases. 4 cases got worse with the skill loaded, and they are included in that figure.
Other measured skills in the registry, with their headline benchmark lift.