Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Create a CHANGELOG.md following keepachangelog.com conventions with version history backfilled from GitHub releases or git tags. Use when the user asks to "create a changelog", "add a changelog", "initialize changelog", "start a changelog", "set up changelog", "generate changelog", or "backfill changelog".
.claude/skills/tobihagemann-create-changelog/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-10 | ✗→✓ | ▲ Improved | -34% | 0% |
| case-12 | ✗→✓ | ▲ Improved | -27% | 0% |
| case-23 | ✗→✓ | ▲ Improved | 195% | 0% |
| case-20 | ✓→✗ | ▼ Worse | 2% | 0% |
| case-04 | ✓→✓ | ＝ Same ✓ | 23% | 0% |
Create a changelog backfilled with version history.
$changelog-rules SkillRun the $changelog-rules skill to load shared changelog conventions.
Use git rev-parse --show-toplevel to find the repository root. Look for the changelog file per $changelog-rules. If it already exists, use request_user_input to confirm overwriting it. If the user declines (chooses the free-form "Other" option or asks to abort), stop without running the remaining steps.
Collect release history from the most authoritative source available:
gh release list --limit 100 --json tagName,name,publishedAt,body to get release notes. For each release, parse the body into changelog entries.git tag --sort=-v:refname to list tags. For each consecutive tag pair, run git log <older-tag>..<newer-tag> --oneline to collect commit summaries.For each version, classify entries into the standard change types and apply the changelog-worthiness criteria per $changelog-rules.
Write the changelog following the $changelog-rules file structure and conventions.
Briefly summarize how many versions were backfilled and which source was used (GitHub releases or git tags).
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 3,017 | 5,742 | +90% | 1 | 1 | 0% | 337 | 609 | +81% | 0 | 0 | — |
case-02 | fail→fail | 3,746 | 5,267 | +41% | 1 | 1 | 0% | 617 | 651 | +6% | 0 | 0 | — |
case-03 | fail→fail | 6,645 | 4,509 | -32% | 1 | 1 | 0% | 386 | 664 | +72% | 0 | 0 | — |
case-04 | pass→pass | 3,208 | 2,022 | -37% | 1 | 1 | 0% | 509 | 626 | +23% | 0 | 0 | — |
case-05 | pass→pass | 5,106 | 2,347 | -54% | 1 | 1 | 0% | 951 | 707 | -26% | 0 | 0 | — |
case-06 | pass→pass | 13,732 | 3,746 | -73% | 1 | 1 | 0% | 2,312 | 970 | -58% | 0 | 0 | — |
case-07 | pass→pass | 5,046 | 3,022 | -40% | 1 | 1 | 0% | 897 | 914 | +2% | 0 | 0 | — |
case-08 | fail→fail | 7,695 | 2,101 | -73% | 1 | 1 | 0% | 1,241 | 675 | -46% | 0 | 0 | — |
case-09 | fail→fail | 11,486 | 2,564 | -78% | 1 | 1 | 0% | 1,743 | 737 | -58% | 0 | 0 | — |
case-10 | fail→pass | 7,122 | 2,239 | -69% | 1 | 1 | 0% | 1,068 | 702 | -34% | 0 | 0 | — |
case-11 | pass→pass | 9,731 | 1,819 | -81% | 1 | 1 | 0% | 1,502 | 601 | -60% | 0 | 0 | — |
case-12 | fail→pass | 6,380 | 2,158 | -66% | 1 | 1 | 0% | 925 | 674 | -27% | 0 | 0 | — |
case-13 | fail→fail | 9,544 | 2,056 | -78% | 1 | 1 | 0% | 1,596 | 664 | -58% | 0 | 0 | — |
case-14 | fail→fail | 7,104 | 2,116 | -70% | 1 | 1 | 0% | 1,180 | 669 | -43% | 0 | 0 | — |
case-15 | pass→pass | 5,131 | 2,427 | -53% | 1 | 1 | 0% | 892 | 724 | -19% | 0 | 0 | — |
case-16 | pass→pass | 5,742 | 1,565 | -73% | 1 | 1 | 0% | 1,029 | 566 | -45% | 0 | 0 | — |
case-17 | pass→pass | 3,349 | 2,876 | -14% | 1 | 1 | 0% | 558 | 839 | +50% | 0 | 0 | — |
case-18 | pass→pass | 7,769 | 4,527 | -42% | 1 | 1 | 0% | 1,339 | 1,084 | -19% | 0 | 0 | — |
case-19 | fail→fail | 11,528 | 5,244 | -55% | 1 | 1 | 0% | 1,836 | 1,145 | -38% | 0 | 0 | — |
case-20 | pass→fail | 5,352 | 9,075 | +70% | 1 | 1 | 0% | 937 | 952 | +2% | 0 | 0 | — |
case-21 | fail→fail | 1,925 | 6,480 | +237% | 1 | 1 | 0% | 222 | 796 | +259% | 0 | 0 | — |
case-22 | fail→fail | 5,959 | 10,931 | +83% | 1 | 1 | 0% | 1,010 | 1,195 | +18% | 0 | 0 | — |
case-23 | fail→pass | 9,999 | 28,905 | +189% | 1 | 1 | 0% | 2,174 | 6,415 | +195% | 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, and 17 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 +9 percentage points is the difference between those two pass rates over the 17 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.