Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Generate release notes for addon packages (non-SDK inference addons, decoder, OCR). Use when preparing release notes or user asks to "generate release notes" for an addon package.
.claude/skills/tetherto-qv-addon-release-notes/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-04 | ✗→✓ | ▲ Improved | 19% | 0% |
| case-11 | ✗→✓ | ▲ Improved | 11% | 0% |
| case-07 | ✗→✓ | ▲ Improved | -15% | 0% |
| case-08 | ✗→✓ | ▲ Improved | 170% | 0% |
| case-09 | ✗→✓ | ▲ Improved | 32% | 0% |
Generate human-readable release notes for addon packages (non-SDK packages like inference addons, decoder, OCR). Collects changes, extracts PRs, validates version bump, and produces a formatted release notes file.
Identify the full set of changes that will land in main for the PR, and validate the version bump in package.json against main:
main. The head is the current branch/commit that the PR will merge.main:main...HEAD (or the PR base commit...HEAD) to list commits and diffs.package.json between main and HEAD.version in package.json is unchanged compared to main, stop and display this exact warning to the user:----------------------------------- WARNING: VERSION BUMP REQUIRED The version in package.json is unchanged compared to main. If this PR includes any changes that must be released in the package, you must bump the package version, commit/push it and re-run this command. -----------------------------------
The goal is to produce a single, combined change set that reflects what will be merged: all commits on the PR branch versus main only.
main...HEAD#123, (#123), or Merge pull request #123gh pr view <number> to get:Create release-notes/vX.Y.Z.md with these guidelines:
# QVAC <package_name_readable> v{VERSION} Release NotesReplace <package_name_readable> with a corresponding name:
decoder-audio: Audio Decoderembed-llamacpp: Embeddings Addonllm-llamacpp: LLM Addontranslation-nmtcpp: NMT Addontts-onnx: TTS ONNX Addontranscription-whispercpp: Transcription Whisper Addonmarkdown ## Pull Requests
release-notes/vX.Y.Z.md:
markdown# QVAC OCR Addon v0.4.0 Release Notes This release introduces automated GitHub releases and improves mobile test reliability. ## Features ### Automated GitHub Releases The release process is now automated with enforced release notes. When a version bump is detected on merge to main, a GitHub release is automatically created using the corresponding release notes file. ## Bug Fixes ### Mobile E2E Test Workflow Fix Fixed an issue where mobile E2E tests would fail when the "On PR Trigger" workflow was manually run via workflow_dispatch. ## Pull Requests - [#67](https://github.com/tetherto/qvac/pull/67) - Fix mobile E2E tests workflow_dispatch - [#70](https://github.com/tetherto/qvac/pull/70) - feat: automate GitHub releases with mandatory release notes
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 6,296 | 6,173 | -2% | 1 | 1 | 0% | 921 | 1,532 | +66% | 0 | 0 | — |
case-12 | pass→pass | 10,116 | 6,035 | -40% | 1 | 1 | 0% | 1,629 | 2,292 | +41% | 0 | 0 | — |
case-02 | fail→fail | 4,515 | 4,719 | +5% | 1 | 1 | 0% | 700 | 1,458 | +108% | 0 | 0 | — |
case-03 | fail→fail | 11,642 | 5,120 | -56% | 1 | 1 | 0% | 1,737 | 1,554 | -11% | 0 | 0 | — |
case-04 | fail→pass | 9,178 | 2,892 | -68% | 1 | 1 | 0% | 1,457 | 1,730 | +19% | 0 | 0 | — |
case-05 | pass→pass | 7,961 | 2,869 | -64% | 1 | 1 | 0% | 1,281 | 1,682 | +31% | 0 | 0 | — |
case-11 | fail→pass | 10,873 | 3,576 | -67% | 1 | 1 | 0% | 1,543 | 1,711 | +11% | 0 | 0 | — |
case-06 | pass→pass | 13,884 | 4,046 | -71% | 1 | 1 | 0% | 2,404 | 1,901 | -21% | 0 | 0 | — |
case-07 | fail→pass | 11,274 | 2,517 | -78% | 1 | 1 | 0% | 1,972 | 1,686 | -15% | 0 | 0 | — |
case-08 | fail→pass | 3,538 | 2,209 | -38% | 1 | 1 | 0% | 609 | 1,643 | +170% | 0 | 0 | — |
case-09 | fail→pass | 5,956 | 2,066 | -65% | 1 | 1 | 0% | 1,155 | 1,530 | +32% | 0 | 0 | — |
case-10 | fail→pass | 9,704 | 4,136 | -57% | 1 | 1 | 0% | 1,490 | 1,937 | +30% | 0 | 0 | — |
case-13 | pass→pass | 8,655 | 2,157 | -75% | 1 | 1 | 0% | 1,420 | 1,579 | +11% | 0 | 0 | — |
case-14 | fail→fail | 7,995 | 1,794 | -78% | 1 | 1 | 0% | 1,207 | 1,472 | +22% | 0 | 0 | — |
case-15 | pass→pass | 6,933 | 2,134 | -69% | 1 | 1 | 0% | 1,114 | 1,493 | +34% | 0 | 0 | — |
case-16 | fail→pass | 7,800 | 2,204 | -72% | 1 | 1 | 0% | 1,389 | 1,534 | +10% | 0 | 0 | — |
case-17 | pass→pass | 12,743 | 3,902 | -69% | 1 | 1 | 0% | 2,097 | 1,962 | -6% | 0 | 0 | — |
case-18 | fail→pass | 5,936 | 2,010 | -66% | 1 | 1 | 0% | 886 | 1,557 | +76% | 0 | 0 | — |
case-19 | fail→pass | 6,080 | 1,263 | -79% | 1 | 1 | 0% | 930 | 1,380 | +48% | 0 | 0 | — |
case-20 | fail→pass | 12,017 | 3,681 | -69% | 1 | 1 | 0% | 1,817 | 1,809 | -0% | 0 | 0 | — |
case-21 | fail→fail | 12,210 | 15,995 | +31% | 1 | 1 | 0% | 2,060 | 3,493 | +70% | 0 | 0 | — |
case-22 | fail→fail | 10,566 | 5,133 | -51% | 1 | 1 | 0% | 1,968 | 2,130 | +8% | 0 | 0 | — |
case-23 | fail→fail | 24,012 | 16,919 | -30% | 1 | 1 | 0% | 5,257 | 4,630 | -12% | 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 21 counted toward the lift figure. The other 2 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 +43 percentage points is the difference between those two pass rates over the 21 comparable cases.
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.