Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Update the Unreleased section of CHANGELOG.md based on current changes. No-op if CHANGELOG.md does not exist. Use when the user asks to "update changelog", "add to changelog", "update the changelog", "changelog entry", "add changelog entry", or "log this change".
.claude/skills/tobihagemann-update-changelog/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-11 | ✗→✓ | ▲ Improved | 202% | 0% |
| case-13 | ✗→✓ | ▲ Improved | 40% | 0% |
| case-06 | ✓→✗ | ▼ Worse | 1% | 0% |
| case-09 | ✓→✗ | ▼ Worse | -4% | 0% |
| case-14 | ✓→✗ | ▼ Worse | 50% | 0% |
Update the Unreleased section of the changelog based on the current changes.
$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 does not exist, skip this skill. Do not create it.
Determine what changed:
git diff --cached for staged changesgit diff for unstaged changesApply the $changelog-rules changelog-worthiness and net-delta criteria. Skip fixes to code introduced by the same branch or PR.
If no changes are changelog-worthy, skip this skill.
Read the current Unreleased section of the changelog. Look for entries that relate to the same feature or fix. This prevents duplicates across multiple commits for the same body of work.
Add or update entries in the Unreleased section following $changelog-rules conventions. Create subsection headers as needed (e.g., ### Added).
Then call update_plan to mark this step completed and continue with the next step of the active workflow.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-10 | fail→fail | 7,155 | 4,225 | -41% | 1 | 1 | 0% | 1,088 | 602 | -45% | 0 | 0 | — |
case-01 | fail→fail | 2,340 | 5,187 | +122% | 1 | 1 | 0% | 316 | 717 | +127% | 0 | 0 | — |
case-02 | fail→fail | 2,381 | 4,928 | +107% | 1 | 1 | 0% | 362 | 600 | +66% | 0 | 0 | — |
case-03 | fail→fail | 3,834 | 4,296 | +12% | 1 | 1 | 0% | 620 | 627 | +1% | 0 | 0 | — |
case-04 | fail→fail | 4,319 | 4,818 | +12% | 1 | 1 | 0% | 231 | 610 | +164% | 0 | 0 | — |
case-05 | fail→fail | 4,235 | 5,672 | +34% | 1 | 1 | 0% | 172 | 659 | +283% | 0 | 0 | — |
case-06 | pass→fail | 4,333 | 5,263 | +21% | 1 | 1 | 0% | 659 | 667 | +1% | 0 | 0 | — |
case-07 | fail→fail | 5,282 | 7,539 | +43% | 1 | 1 | 0% | 773 | 946 | +22% | 0 | 0 | — |
case-08 | pass→pass | 7,245 | 4,465 | -38% | 1 | 1 | 0% | 1,141 | 1,086 | -5% | 0 | 0 | — |
case-09 | pass→fail | 3,962 | 5,899 | +49% | 1 | 1 | 0% | 685 | 656 | -4% | 0 | 0 | — |
case-11 | fail→pass | 3,854 | 7,824 | +103% | 1 | 1 | 0% | 371 | 1,119 | +202% | 0 | 0 | — |
case-12 | pass→pass | 3,777 | 12,363 | +227% | 1 | 1 | 0% | 559 | 1,646 | +194% | 0 | 0 | — |
case-13 | fail→pass | 5,809 | 7,507 | +29% | 1 | 1 | 0% | 841 | 1,179 | +40% | 0 | 0 | — |
case-14 | pass→fail | 4,369 | 7,945 | +82% | 1 | 1 | 0% | 810 | 1,211 | +50% | 0 | 0 | — |
case-15 | pass→pass | 7,553 | 7,115 | -6% | 1 | 1 | 0% | 1,334 | 1,716 | +29% | 0 | 0 | — |
case-16 | fail→fail | 5,120 | 5,388 | +5% | 1 | 1 | 0% | 291 | 836 | +187% | 0 | 0 | — |
case-17 | pass→fail | 3,847 | 4,875 | +27% | 1 | 1 | 0% | 597 | 755 | +26% | 0 | 0 | — |
case-18 | pass→fail | 3,259 | 4,642 | +42% | 1 | 1 | 0% | 483 | 652 | +35% | 0 | 0 | — |
case-19 | fail→fail | 10,967 | 2,784 | -75% | 1 | 1 | 0% | 1,896 | 804 | -58% | 0 | 0 | — |
case-20 | pass→fail | 3,219 | 4,853 | +51% | 1 | 1 | 0% | 413 | 626 | +52% | 0 | 0 | — |
case-21 | pass→fail | 5,070 | 6,059 | +20% | 1 | 1 | 0% | 728 | 629 | -14% | 0 | 0 | — |
case-22 | pass→fail | 3,044 | 5,937 | +95% | 1 | 1 | 0% | 423 | 674 | +59% | 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 6 counted toward the lift figure. The other 16 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 -27 percentage points is the difference between those two pass rates over the 6 comparable cases. 8 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.