asc-whats-new-writer
Generate engaging, localized App Store release notes (What's New) from git log, bullet points, or free text using canonical metadata under `./metadata`. Optionally pairs with promotional text updates.
- License 许可
- MIT
- 最近更新
- 2026-11-13
- Stars
- ★ 5
- Forks
- 0
01 前言 Front Matter
- name
- asc-whats-new-writer
- description
- Generate engaging, localized App Store release notes (What's New) from git log, bullet points, or free text using canonical metadata under `./metadata`. Optionally pairs with promotional text updates.
02 资源文件 3
- 许可证LICENSE1.0 KB
- 参考文档references/release_notes_guidelines.md5.0 KB
- 参考文档UPSTREAM.json1.6 KB
03 SKILL.md 正文
asc What's New Writer
Generate engaging, localized release notes from flexible input. Optionally pair with promotional text updates.
Preconditions
- Metadata pulled locally into canonical files via
asc metadata pull --app "APP_ID" --version "1.2.3" --dir "./metadata". OR: user provides keywords manually. - Auth configured for upload (
asc auth loginorASC_*env vars). - The primary locale is
en-USunless the user specifies otherwise.
Before You Start
- Read
references/release_notes_guidelines.mdfor tone, structure, and examples. - Identify the latest version directory under
metadata/version/(highest semver). Use this for all metadata reads. - Enumerate existing locales by listing the JSON files in that version directory.
Phase 1: Gather Input
Accept one of three input modes (auto-detect):
Git Log
Parse commits since the last tag:
# Find latest tag
git describe --tags --abbrev=0
# List commits since that tag
git log $(git describe --tags --abbrev=0)..HEAD --oneline --no-merges
Filter out noise: merge commits, dependency bumps, CI changes, formatting-only commits. Extract user-facing changes.
Bullet Points
User provides rough bullets like:
- "improved search"
- "fixed crash on launch"
- "added sleep timer"
Free Text
User describes changes conversationally:
"We made search faster, fixed that annoying crash when you open the app, and added a sleep timer feature"
The skill extracts and structures the changes from the text.
No Input Provided
Prompt the user: "What changed in this release? You can paste git log output, bullet points, or just describe the changes."
Phase 2: Draft Notes (Primary Locale)
Step 1: Classify Changes
Group changes into sections per the guidelines:
- New — new features or capabilities
- Improved — enhancements to existing features
- Fixed — bug fixes users would notice
Omit empty sections. If all changes are fixes, only show "Fixed."
Step 2: Write Benefit-Focused Copy
Follow the tone rules from references/release_notes_guidelines.md:
- Describe user impact, not implementation details
- Use direct address ("you") and action verbs
- Be specific — mention concrete improvements
Step 3: Front-Load the Hook
The first ~170 characters are the only visible part before "more." Lead with the single most impactful change in a complete, compelling sentence.
Step 4: Echo Keywords for Conversion
- Read
keywordsfrommetadata/version/{latest}/{primary-locale}.json- These canonical files are also what
asc metadata keywords ...reads and writes.
- These canonical files are also what
- If the field is empty or missing, skip this step
- Identify keywords relevant to the changes being described
- Weave them naturally into the notes — never force or stuff
Step 5: Respect Character Limits
- Keep total length between 500-1500 characters in the primary locale
- This leaves room for localized expansions (some languages expand 30-40%)
- Hard limit: 4,000 characters
Step 6: Optionally Draft Promotional Text
If the user wants it, draft a 170-char promotional text that:
- Summarizes the update's theme in one punchy line
- Can reference seasonal events
- Is updatable without a new submission
Present Draft
Show the draft to the user with character count. Wait for approval before localizing.
Phase 3: Localize
Translate the approved notes to all existing locales.
Translation Rules
- Use formal register and formal "you" forms (Russian: вы, German: Sie, French: vous, Spanish: usted, Dutch: u, Italian: Lei)
- Adapt tone to local market — playful English may need adjustment for formal markets (ja, de-DE)
- Do NOT literally translate idioms — adapt them to local equivalents
- A playful tone in English may need to be more respectful or formal in other cultures
Locale-Specific Keyword Echo
For each locale:
- Read
keywordsfrommetadata/version/{latest}/{locale}.json - Echo locale-specific keywords naturally in the translated notes
- If keywords field is empty, skip echo for that locale
Validate
- All translations must be ≤ 4,000 characters
- Promotional text must be ≤ 170 characters per locale
- If a translation exceeds the limit, shorten it — never truncate mid-sentence
Phase 4: Review & Upload
Step 1: Present Summary
Show a table of all locales with their notes and character counts:
| Locale | What's New (first 80 chars...) | Chars | Promo Text | Chars |
|--------|-------------------------------|-------|------------|-------|
| en-US | Search just got faster — ... | 847 | New sleep… | 142 |
| ar-SA | البحث أصبح أسرع — ... | 923 | نوم جديد… | 138 |
| ... | ... | ... | ... | ... |
Step 2: Wait for Approval
Do not upload without user confirmation.
Step 3: Upload
Upload via asc (verify exact syntax with asc --help):
# Individual locale direct update
asc apps info edit --app "APP_ID" --version-id "VERSION_ID" --locale "en-US" --whats-new "Your release notes here"
# Bulk canonical-metadata push after writing ./metadata/version/<version>/<locale>.json
asc metadata push --app "APP_ID" --version "1.2.3" --dir "./metadata" --dry-run
asc metadata push --app "APP_ID" --version "1.2.3" --dir "./metadata"
If promotional text was drafted, either include --promotional-text "..." in the direct update command or write promotionalText into the canonical JSON before asc metadata push.
Step 4: Handle Failures
On partial upload failure:
- Report which locales succeeded and which failed
- Offer to retry failed locales
Metadata File Paths
- Keywords:
metadata/version/{latest-version}/{locale}.json→keywordsfield - Current What's New:
metadata/version/{latest-version}/{locale}.json→whatsNewfield - Latest version: highest semver directory under
metadata/version/ - The canonical
./metadatatree is whatasc metadata pull,asc metadata push, andasc metadata keywords ...operate on. - Follows the same metadata resolution conventions as
asc-aso-audit
Notes
- What's New is not indexed for App Store search — write for humans, not algorithms.
- Promotional text is the only metadata field updatable without a new submission.
- The 170-char visible window is the most important part of your release notes.
- Each app update triggers algorithm re-evaluation — the act of updating matters, even if the text doesn't affect ranking.
- Ideal update cadence: every 2-4 weeks.
- For full metadata translation (all fields), use
asc-localize-metadatainstead. - For keyword research and optimization, use
asc-aso-auditfirst. - If the local keyword field is stale before drafting, refresh it with
asc metadata pullor inspect planned keyword changes withasc metadata keywords diff.
04 同仓库更多 Skills
icons-pro-max
A portable icon catalog and rendering guide for the current project: payment methods, QR codes, social/app brand marks, download badges, the lucide-react UI icon set, 145 tech-stack logos, 86 full-color AI provider marks, 516 Claude connector icons, and 25 Microsoft app logos. Gives the exact asset path, React component, canonical size, brand color, render recipe, and accessibility rule for each icon — plus an anti-slop checklist so an agent never redraws a logo, recolors a brand mark, distorts an aspect ratio, or re-inlines a duplicated icon. Use whenever adding, rendering, swapping, or auditing an icon anywhere in a website, app, or document.
design-spacing-rhythm
The most complete spacing & alignment skill for AI-generated UI and written documents. Reads the layout, sets three dials (SPACING_STEP / DENSITY / ALIGNMENT_RIGOR), derives ONE scale, and lets proximity carry meaning — establishing a spacing scale, vertical rhythm, optical alignment, whitespace hierarchy, responsive/fluid density, accessibility floors, component recipes, document & long-form (Word / Docs / PDF / Markdown) spacing, and an anti-slop checklist. Every rule is CONTEXTUAL: read the situation, then pull only what fits. Not a blanket "add more padding" pass.
apple
For Apple app UI and code (SwiftUI, Xcode, HIG) and scroll scenes. Use when a question or task needs HIG reviews, App Store Connect (app records, TestFlight, signing, notarizing, ad hoc, What's New, ASO, Apple Ads, Wall of Apps, screenshots), SwiftData, Swift 6, Swift Testing, Core Data, Xcode builds, disk space, Tuist apps or pinned or scrubbed web sections. Not for launch/preview videos, captioned screenshots, Play Store art, GSAP, web springs (apple-design) or Swift questions (write-swift).
xcode-disk-cleanup
Audit and safely clean Xcode-related developer storage on macOS. Use whenever a developer mentions low disk space, Xcode storage, DerivedData, simulator devices or runtimes, stale beta platforms in Xcode Settings, DeviceSupport, archives, dSYMs, CocoaPods caches, simulator dyld caches, project .build folders, downloaded Xcode DMGs/XIPs, duplicate Xcode installers, or old Xcode applications. Also use this skill proactively when you hit low storage during other work, such as a build, download, or install failing with an out-of-disk-space error. Always measure first, report recoverable GiB with evidence, and obtain explicit itemized approval before any mutation.
xcode-project-analyzer
Audit Xcode project configuration, build settings, scheme behavior, and script phases to find build-time improvements with explicit approval gates. Use when a developer wants project-level build analysis, slow incremental builds, guidance on target dependencies, build settings review, run script phase analysis, parallelization improvements, or module-map and DEFINES_MODULE configuration.
xcode-compilation-analyzer
Analyze Swift and mixed-language compile hotspots using build timing summaries and Swift frontend diagnostics, then produce a recommend-first source-level optimization plan. Use when a developer reports slow compilation, type-checking warnings, expensive clean-build compile phases, long CompileSwiftSources tasks, warn-long-function-bodies output, or wants to speed up Swift type checking.