Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Create release notes and product updates that developers actually read and care about. This skill covers changelog formatting, versioning communication, breaking change announcements, deprecation notices, and building anticipation for new features. Trigger phrases: "changelog",...
.claude/skills/sickn33-changelog-updates/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flash | 100% | 2 |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-06 | ✗→✓ | ▲ Improved | 189% | 0% |
| case-07 | ✗→✓ | ▲ Improved | 181% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 167% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 173% | 0% |
| case-15 | ✗→✓ | ▲ Improved | 140% | 0% |
Read the detailed guide before executing this skill. It retains the complete procedure and reference material. Treat its safety, prerequisites, and validation requirements as mandatory. For focused work, load the relevant sections; for end-to-end work, read the guide completely.
Use this skill when you need create release notes and product updates that developers actually read and care about. This skill covers changelog formatting, versioning communication, breaking change announcements, deprecation notices, and building anticipation for new features. Trigger phrases: "changelog",...
Release notes are developer communication, not documentation. When done well, they build trust, demonstrate momentum, and turn updates into marketing moments.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | pass→pass | 5,392 | 5,172 | -4% | 1 | 1 | 0% | 1,107 | 4,831 | +336% | 0 | 0 | — |
case-02 | fail→fail | 7,043 | 6,367 | -10% | 1 | 1 | 0% | 1,289 | 5,019 | +289% | 0 | 0 | — |
case-03 | pass→pass | 7,577 | 5,799 | -23% | 1 | 1 | 0% | 1,421 | 4,948 | +248% | 0 | 0 | — |
case-04 | pass→pass | 12,736 | 12,727 | -0% | 1 | 1 | 0% | 2,232 | 6,272 | +181% | 0 | 0 | — |
case-05 | pass→pass | 9,340 | 7,608 | -19% | 1 | 1 | 0% | 1,923 | 5,377 | +180% | 0 | 0 | — |
case-06 | fail→pass | 11,242 | 8,355 | -26% | 1 | 1 | 0% | 1,935 | 5,585 | +189% | 0 | 0 | — |
case-07 | fail→pass | 10,456 | 10,200 | -2% | 1 | 1 | 0% | 2,126 | 5,980 | +181% | 0 | 0 | — |
case-08 | fail→pass | 14,258 | 17,613 | +24% | 1 | 1 | 0% | 2,733 | 7,300 | +167% | 0 | 0 | — |
case-09 | fail→pass | 9,706 | 6,805 | -30% | 1 | 1 | 0% | 2,044 | 5,579 | +173% | 0 | 0 | — |
case-10 | pass→pass | 6,266 | 5,808 | -7% | 1 | 1 | 0% | 1,361 | 5,231 | +284% | 0 | 0 | — |
case-11 | pass→pass | 12,472 | 12,651 | +1% | 1 | 1 | 0% | 2,144 | 6,245 | +191% | 0 | 0 | — |
case-12 | pass→pass | 8,143 | 8,038 | -1% | 1 | 1 | 0% | 1,480 | 5,337 | +261% | 0 | 0 | — |
case-13 | pass→pass | 10,256 | 7,794 | -24% | 1 | 1 | 0% | 1,735 | 5,380 | +210% | 0 | 0 | — |
case-14 | pass→pass | 16,738 | 15,374 | -8% | 1 | 1 | 0% | 3,002 | 6,714 | +124% | 0 | 0 | — |
case-15 | fail→pass | 17,133 | 17,477 | +2% | 1 | 1 | 0% | 2,981 | 7,141 | +140% | 0 | 0 | — |
case-16 | fail→fail | 2,973 | 4,706 | +58% | 1 | 1 | 0% | 502 | 4,752 | +847% | 0 | 0 | — |
case-17 | fail→pass | 5,763 | 6,597 | +14% | 1 | 1 | 0% | 1,139 | 5,212 | +358% | 0 | 0 | — |
case-18 | fail→pass | 3,187 | 3,213 | +1% | 1 | 1 | 0% | 579 | 4,501 | +677% | 0 | 0 | — |
case-19 | pass→pass | 7,706 | 5,671 | -26% | 1 | 1 | 0% | 1,717 | 5,299 | +209% | 0 | 0 | — |
case-20 | pass→pass | 17,901 | 17,847 | -0% | 1 | 1 | 0% | 4,369 | 8,370 | +92% | 0 | 0 | — |
case-21 | pass→pass | 5,591 | 6,331 | +13% | 1 | 1 | 0% | 991 | 4,993 | +404% | 0 | 0 | — |
case-22 | pass→pass | 14,227 | 12,955 | -9% | 1 | 1 | 0% | 2,392 | 6,113 | +156% | 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. The headline lift of +32 percentage points is the difference between those two pass rates over the 22 comparable cases.
The publisher has shipped newer versions since this run, so these numbers describe v1, not the version currently listed.
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.