agentleFS
cenconq25cenconq25/claude-code-app-studio/.claude/skills/changelog/SKILL.md

changelog

Skill40 starsChanged 7 months agoMITView on GitHub

Auto-generate a changelog from git commits, sprint data, and design docs. Produces both internal (engineering-facing) and user-facing (release-notes) versions for a given release window.

160 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use playtest-report, team-audio and 2 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • CoralGameSkill

    playtest-report

    Generates a structured playtest report template or analyzes existing playtest notes into a structured format. Use this to standardize playtest feedback collection and analysis.

    Also in 12 projects that use this

  • CoralGameSkill

    team-audio

    Orchestrate audio team: audio-director + sound-designer + technical-artist + gameplay-programmer for full audio pipeline from direction to implementation.

    Also in 12 projects that use this

  • +2 more from CoralGame

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 4 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 149 edited versions we read keep these.

  • 4

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Default: --from = previous release tag, --to = HEAD.

    line 27
  2. Distinctive:

    Run each commit through a classifier:

    line 57
  3. Very specific:

    Ask before writing to production/releases/changelog-internal-[version].md.

    line 119
  4. Distinctive:

    Keep entries terse — one line per change unless context truly needs two.

    line 136
  5. Very specific:

    If a translation pipeline exists (vendor, MT), tag the user-facing copy as # UNTRANSLATED and write it to …

    line 184
  6. Distinctive:

    If a vendor is not configured, surface this as an open item to resolve before submission.

    line 188
  7. Distinctive:

    Ask before each write.

    line 191

“Common” is measured: kept by at least 6 in 10 of the 149 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Changelog

Translate a git history window plus the corresponding sprint and PRD context into a clean changelog. The internal version explains what shipped and why; the user-facing version is short, plain, and respectful of store character limits.


Phase 1: Resolve the Window

Parse arguments:

  • --from=<tag-or-date> and --to=<tag-or-date> define the explicit window.
  • --version=<v> looks up the previous release tag and uses it as the --from.
  • Default: --from = previous release tag, --to = HEAD.

Use Bash:

git tag --sort=-creatordate | head -n 5
git log --pretty=format:'%h%x09%s%x09%an' [from]..[to]

Capture commits, authors, and SHAs.


Phase 2: Pull Auxiliary Context

Read in parallel:

  • Sprint files closed within the window: production/sprints/sprint-*.md.
  • Story files closed within the window: production/epics/**/story-*.md filtered by Status: Complete and Closed: [in-window].
  • Bug records closed within the window: production/qa/bugs/BUG-*.md filtered by Status: Closed and a closing date in window.
  • PRD updates in the window via git log -- design/prd/.
  • ADRs accepted in the window via git log -- docs/architecture/.

Phase 3: Categorize Commits

Run each commit through a classifier:

  • Features — commit message includes "feat" / "feature" / refers to a story closing as new functionality.
  • Improvements — "perf", "improve", "polish", "tune".
  • Fixes — "fix", "bug", references a BUG ID.
  • Security — "security", "CVE", references a security audit finding.
  • Internal — "refactor", "chore", "test", "ci", "docs", "style".
  • Reverted — Revert "..." style messages.

For each commit, attempt to enrich:

  • Linked story (commit body or branch name often contains STORY-NN).
  • Linked bug (BUG-NN).
  • Linked PRD section.

Phase 4: Internal Changelog

Render at one entry per change. Structure:

# Internal Changelog — [version]

Window: [from] -> [to]
Author count: [N]
Commit count: [N]

## Features
- [STORY-NN] [headline] — [PRD ref] (commits: abc123, def456)
  Notes: [summary, framework notes, ADR ref]

## Improvements
- [headline] — [perf-profile reference if perf-related]

## Fixes
- [BUG-NN] [headline] (commits: ...)

## Security
- [security-audit ref] [summary]

## Internal / Refactors
- [headline]

## Reverted
- [reverted commit + reason if known]

## Notes for Engineers
- API changes: [list]
- Migration steps: [list]
- Deprecations introduced: [list]
- New ADRs: [IDs]

## Stats
- Stories closed: [N]
- Bugs closed: [N]
- Test files added: [N]
- LOC delta: [+P / -Q]

Ask before writing to production/releases/changelog-internal-[version].md.


Phase 5: User-Facing Changelog

Translate Internal -> User-facing. Apply these rules:

  • Drop Internal / Refactor entries entirely.
  • Drop Reverted entries unless user-visible.
  • Rewrite headlines to user benefit, not implementation:
    • "Refactor cart reducer to immutable state" -> drop.
    • "Add Apple Pay to checkout" -> "Apple Pay now works at checkout."
    • "Fix race condition in token refresh" -> "Sign-in is more reliable when switching networks."
  • Group by theme that makes sense to a user, not an engineer.
  • Keep entries terse — one line per change unless context truly needs two.

Render two formats:

Long form (in-app updates feed, release notes page)

# What's New — [version]

## New
- Apple Pay at checkout
- Dark mode for the home feed

## Improvements
- Faster cold start
- Smoother list scrolling

## Fixes
- Sign-in is more reliable when switching networks
- The paywall no longer flickers on slow connections

## Coming soon
- [list teased items if approved by marketing]

Short form (App Store / Play Store What's New)

App Store limit: 4000 chars. Play Store limit: 500 chars.

Produce a 500-char version that hits the highlights:

We've added Apple Pay at checkout and dark mode for the home feed.
Cold start is faster and list scrolling is smoother. Sign-in is more
reliable when you switch networks, and we squashed a paywall flicker.
Thanks for using [App Name].

Show character counts and ask the user to approve.


Phase 6: Translate User-Facing for Locales

Pull the target locale list from .claude/docs/technical-preferences.md. For each locale:

  • If a translation pipeline exists (vendor, MT), tag the user-facing copy as # UNTRANSLATED and write it to production/releases/changelog-user-[version]-[locale].md for the vendor.
  • If a vendor is not configured, surface this as an open item to resolve before submission.

Ask before each write.


Phase 7: Render and Save

Ask:

  • "May I write Internal changelog to production/releases/changelog-internal-[version].md?"
  • "May I write User long-form to production/releases/changelog-user-[version].md?"
  • "May I write Short form (App Store) to production/releases/whats-new-[version]-app-store.md?"
  • "May I write Short form (Play Store) to production/releases/whats-new-[version]-play-store.md?"

Phase 8: Update State

Append to production/session-state/active.md:

## Changelog — [date]
- Window: [from] -> [to]
- Internal: [path]
- User long: [path]
- User App Store: [path]
- User Play Store: [path]
- Locales pending translation: [list]
- Next: /patch-notes for short-form polishing, then /release-checklist

Quality Gates / PASS-FAIL

  • PASS — internal changelog has every commit categorized; user-facing drops internal noise; short-form fits within store limits; no PRD-private detail leaks to user-facing.
  • FAIL — internal text leaked into user-facing, or short-form exceeds store limits.

Examples

Example 1 — minor v1.3.0: Window covers 2 sprints, 47 commits, 12 stories closed. Internal changelog produced; user-facing trimmed to 8 bullets; short-form 320 chars. 3 locales tagged for translation.

Example 2 — patch v1.2.1 (hotfix): 1 commit, 1 BUG. Internal: one fix entry. User-facing: one bullet ("We fixed an issue causing sign-in to fail intermittently."). Short- form: 95 chars.


Next Steps

  • /patch-notes to polish the short-form for store submission.
  • /release-checklist to make sure submission metadata includes the user-facing text.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.