agentleFS
1-Pankaj1-Pankaj/AutonomousMarketing/.agents/skills/apple-design-hig/AGENTS.md

AutonomousMarketing / apple-design-hig

AGENTS.md1 starsChanged 12 days agoMITView on GitHub

Operating manual for AI agents working in this repository, or using it from another project. Apple Design Skill: a design-review skill (SKILL.md) backed by 123 pages of Apple's Human Interface Guidelines pulled from developer.apple.com into references/hig/, one curated guide, a generated lookup table, and the script that regenerates all of it. 1. Read SKILL.md in full. It defines the reviewer's stance, Apple's eight design principles, the five audit lenses, the report format, and the improvement process. 2. Pick files from…

1 other project uses a version of this.

Adopted by

Most recently taken up by

  • Other versions1 project uses a version of this.

    Word for word · 1

    1. dickwu/apple-design-skill★ 2.6k100% same

    Edited · 0

    None found.

    Wrote their own version · 0

    None found.

    How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.

  • Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.

    Of the AGENTS.md files we read, the 31,439 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.

  • In this repository18 other files it gives its agents.

The file

We split this AGENTS.md into the rules it gives an agent

7 rules. About 0 are common, 1 is distinctive, and 6 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 6

    Very specific

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

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

    Read SKILL.md in full. It defines the reviewer's stance, Apple's eight design principles, the five audit lenses, the report format, and the …

    line 13
  2. Very specific:

    Write the report in the SKILL.md format: Summary, Critical, Improvements, Craft notes, What works, Platform notes.

    line 25
  3. Very specific:

    After running it: review git diff --stat, spot-check a changed file, run the checks below, refresh liquid-glass.md if a page it links to …

    line 73
  4. Very specific:

    Keep SKILL.md under about 450 lines.

    line 96
  5. Very specific:

    Don't add pages to references/hig/ by hand.

    line 102
  6. Distinctive:

    Commit generated changes separately from hand-written changes.

    line 107
  7. Very specific:

    Do not commit .cache/, .omc/, or other local state.

    line 108

≈ “Common” is an estimate: at least 1 in 5 of the 40,246 AGENTS.md files we read have a rule of the same kind, not the same words. “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
AGENTS.md

AGENTS.md

Operating manual for AI agents working in this repository, or using it from another project.

What this repository is

Apple Design Skill: a design-review skill (SKILL.md) backed by 123 pages of Apple's Human Interface Guidelines pulled from developer.apple.com into references/hig/, one curated guide, a generated lookup table, and the script that regenerates all of it.

Running a design review

  1. Read SKILL.md in full. It defines the reviewer's stance, Apple's eight design principles, the five audit lenses, the report format, and the improvement process.
  2. Pick files from SKILL.md: the always-load set, the "Load by what is on screen" table, and the index of every page by Apple's section. references/hig-lookup.md adds Apple's one-line summary of each page and the list of pages left out; open it to check what a page covers when its name is ambiguous.
  3. Load the always-load set (accessibility.md, layout.md, typography.md, color.md, plus designing-for-ios.md or designing-for-macos.md, plus references/cross-platform.md unless the app is SwiftUI, UIKit, or AppKit) and 3 to 6 more files for what is on screen. Never load the whole directory.
  4. Quote the guideline and cite it as file.md › Heading. Use the user's framework vocabulary (Flutter, React Native, Tauri, Electron, SwiftUI, UIKit, AppKit).
  5. Write the report in the SKILL.md format: Summary, Critical, Improvements, Craft notes, What works, Platform notes. Each finding has a What, a Why with a citation, and a Fix.

File map

PathRoleEdit by hand?
SKILL.mdThe skill: stance, principles, lenses, report format, improvement modeYes
AGENTS.mdThis fileYes
README.mdInstallation and overview for peopleYes
.cursorrulesEntry point for CursorYes
GEMINI.mdEntry point for AntigravityYes
.github/FUNDING.ymlGitHub Sponsors button configurationYes
references/hig-lookup.mdGenerated table of every page with Apple's summary and change date, plus an omitted-pages listNo, re-run the script
references/hig/*.md123 generated pages in Apple's wordingNo, re-run the script
references/hig/liquid-glass.mdCurated Liquid Glass guide; the script never touches itYes
references/cross-platform.mdApple's vocabulary and conventions for Flutter, React Native, Tauri, and ElectronYes
references/code-audit.mdCode patterns to search for (Apple Pay, Liquid Glass, iPhone Duo), with citations and replacementsYes
scripts/pull-hig.mjsPull script, Node 18 or newer, no dependenciesYes

Refreshing the references

node scripts/pull-hig.mjs                 # pull from developer.apple.com into references/
node scripts/pull-hig.mjs --cache .cache  # also keep the raw JSON for fast re-runs

Options: --out <dir>, --lookup <file>, --cache <dir>, --concurrency <n>, --no-prune, --force-prune.

What the script does:

  • Crawls Apple's six HIG sections and their sub-collections (158 pages as of October 2026).
  • Keeps every page that applies to iOS, iPadOS, or macOS, omits Apple-only hardware and services listed in OMITTED_PAGES, and records every omission with a reason in hig-lookup.md.
  • Renders Apple's JSON to Markdown faithfully: headings, paragraphs, lists, tables, notes, and tabbed content as sub-headings. Images and videos are dropped except for check-mark glyphs in comparison tables.
  • Relabels platform headings by device class (Phone (iOS), Tablet (iPadOS), Mobile (iOS, iPadOS), Desktop (macOS), and combinations), drops sections that apply only to tvOS, visionOS, or watchOS (sentences that mention them stay), and rewrites cross-references as local links whose anchors match the relabeled headings. Links to headings that were dropped or that Apple no longer publishes point at the file instead.
  • Keeps each page's change log and turns Apple's "Related" links into a "Related guidelines" list.
  • Prunes generated files for pages Apple removed. Curated files are never pruned, and a run that would prune more than a tenth of the files stops and asks for --force-prune.

After running it: review git diff --stat, spot-check a changed file, run the checks below, refresh liquid-glass.md if a page it links to changed, and, if pages were added or removed, update the page counts in README.md, SKILL.md, and this file, and add or remove the page in the "Every page, by Apple's section" table in SKILL.md. A new "Designing for" page has no platform metadata: if its Platforms covered line doesn't name a platform, add its slug to DESIGNING_FOR_COVERAGE in the script.

Checks before finishing any change

node --check scripts/pull-hig.mjs
node scripts/pull-hig.mjs --cache .cache && git status --short   # second run must change nothing
grep -rl 'doc://' references/hig                                  # must print nothing
npx skills add . --list                                           # must find the apple-design skill
ls references/hig | while read -r f; do sed -n '/^### Every page/,/^## /p' SKILL.md | grep -q "\`$f\`" || echo "not in the page table: $f"; done   # must print nothing
grep -o '`[a-z0-9-]*\.md`' SKILL.md | tr -d '`' | sort -u | while read -r f; do [ -f "references/hig/$f" ] || echo "no such page: $f"; done   # must print nothing

The last two keep SKILL.md and references/hig/ in step in both directions: every page is in the skill's "Every page, by Apple's section" table, and every page the skill names exists.

Conventions

  • Keep SKILL.md under about 450 lines; Anthropic's skill guidance caps the body at 500. It loads in full every time the skill triggers, and it names every guideline page so that no reference sits more than one level away from it.
  • Apple's text in references/hig/ is quoted, not paraphrased. Cross-platform translation lives in references/cross-platform.md, SKILL.md, and liquid-glass.md, and code patterns in references/code-audit.md, not in the pulled files.
  • Don't add pages to references/hig/ by hand. Change OMITTED_PAGES or the platform rule in the script instead.
  • A hand-written file in references/hig/ must be registered in CURATED_FILES and must not start with the generated header (a > Source: <...> line followed by > Section:). Open it with > Curated guide as liquid-glass.md does, so prune can never mistake it for output.
  • Commit generated changes separately from hand-written changes.
  • Do not commit .cache/, .omc/, or other local state.

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.