Detection improvements:
- Remove SAFE_TAGS from glow check (buttons/links with glows are valid)
- Add gradient color parsing (parseGradientColors) for AI palette
detection on gradient backgrounds including buttons
- Detect cyan neon text on dark backgrounds as AI palette
- Resolve gradient backgrounds as dark for glow detection
- Fix pure-black false positive on semi-transparent overlays (a >= 0.9)
- Skip low-contrast/gray-on-color when background is a gradient
- Fix "Only font:" double-colon in browser labels
Test performance:
- Split jsdom fixture tests to Node's test runner (bun + jsdom hangs
after ~13 instances due to resource leak)
- bun test for unit/regex/CLI tests (94 tests, 4s)
- node --test for jsdom fixtures (15 tests, 1.3s)
- Total: 109 tests in ~5s (was 280s+)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Extract all regex-on-HTML checks into shared checkHtmlPatterns()
function called by both browser and Node paths. Eliminates drift
between checkTypography/checkPageTypography and removes separate
checkPageMotion/checkPageGlow functions.
Add parseGradientColors() utility and checkElementGradientDOM()
to detect purple/violet gradient backgrounds on any element
including buttons (bypasses SAFE_TAGS). Fix false low-contrast
findings on gradient backgrounds by returning null from
resolveBackground when a gradient is encountered.
Fix "Only font:" double-colon in browser labels.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
New detections:
- bounce-easing: flags bounce/elastic animation names, animate-bounce
(Tailwind), and cubic-bezier curves with overshoot (y values outside
[0, 1])
- layout-transition: flags explicit transition of width, height, padding,
margin, and max-height/min-width variants; skips transition: all
- dark-glow: flags colored box-shadow with blur > 4px on dark backgrounds
(luminance < 0.1); skips gray shadows, focus rings (no blur), and
non-dark backgrounds
Includes 48 new tests across unit, regex, and jsdom fixture tests with
dedicated should-flag and should-pass HTML fixtures for both categories.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Combine detect-antipatterns-core.mjs, detect-antipatterns.mjs, and
detect-antipatterns-browser-wrapper.js into a single universal file
that auto-detects browser vs Node via IS_BROWSER. Shared constants,
color utilities, and pure detection logic exist once instead of
being duplicated across files.
Build script simplified to strip @browser-strip-start/end markers,
set IS_BROWSER=true, and wrap in IIFE.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The generated browser detector now lives alongside the CLI script in
.claude/skills/critique/scripts/ — clearly a build artifact, not a
hand-maintained source file in public/js/.
- build-browser-detector.js outputs to .claude/ instead of public/js/
- Dev server serves .claude/skills/* for local testing
- All fixture and antipattern-example HTML files updated to new path
- Puppeteer detectUrl reads browser script from same directory
- Browser parity test server updated to serve from .claude/
- Deleted public/js/detect-antipatterns-browser.js
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
DRY refactor:
- detect-antipatterns-core.mjs (297 lines): shared constants (SAFE_TAGS,
OVERUSED_FONTS, GENERIC_FONTS, ANTIPATTERNS), color utilities (parseRgb,
relativeLuminance, contrastRatio, hasChroma, getHue, colorToHex,
isNeutralColor), and pure detection functions (checkBorders, checkColors,
isCardLikeFromProps).
- CLI (889 lines, was 1212): imports from core, keeps jsdom-specific
resolveBackground, page-level analyzers, regex fallback, and CLI logic.
- Browser wrapper (335 lines): template with browser-specific DOM adapters,
highlighting, scan loop. Core is injected at build time.
- build-browser-detector.js: reads core, strips exports, injects into
wrapper, writes to public/js/detect-antipatterns-browser.js (generated).
- Build step added to scripts/build.js (runs before Bun bundling).
Source of truth for detection logic is now the core module. Browser script
is generated — do not edit public/js/detect-antipatterns-browser.js directly.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Major cleanup:
- detectUrl() now injects the browser script via page.evaluate() and
calls window.impeccableScan() instead of reimplementing all detection
logic inline. Removes ~80 lines of triple-duplicated code.
- Removed dead isPureBlackOrWhite function.
- CLI reduced from 1286 to 1212 lines.
New: Puppeteer-powered browser parity tests (detect-antipatterns-browser.test.js):
- Starts a local HTTP server for fixtures
- Loads fixture pages in headless Chrome
- Runs the browser detection script via impeccableScan()
- Verifies findings match expectations for all fixture categories:
borders, colors, layout, typography, partials
8 new browser tests catch desync between CLI and browser script
(like the WeakSet iteration bugs we hit earlier).
puppeteer added as devDependency.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Three fixes:
- Only flag innermost nested cards: if L1>L2>L3, only L3 gets flagged
(not L2). Uses ancestor-filtering after collection pass.
- Lower text threshold from 20 to 10 chars to catch short card content
like "Inner card via CSS."
- isCardLike now also checks raw inline style attribute for box-shadow
and border-radius (jsdom doesn't resolve CSS shorthands). Tightened
heuristic: shadow or border is mandatory (not optional).
Fixes false positive on layout-should-pass where a tinted subsection
(rounded + bg, no shadow) inside a card was incorrectly flagged.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Nested cards: fixed dedup to use WeakSet on actual elements instead
of tag-name key, so all nested card instances are found (not just
the first div-in-div pair). Now catches all 4+ nesting examples.
- Dropped identical-card-grid: too many legitimate uses (data displays,
pricing cards, navigation tiles) make false positives unavoidable.
- Removed from CLI, browser script, tests, and ANTIPATTERNS registry.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Four new layout detections:
- nested-cards: jsdom DOM walk finds card-like elements (shadow + rounded + bg)
nested inside other card-like elements. Excludes dropdowns (absolute/fixed),
form inputs, code blocks, badges (<20 chars), and known component classes.
- identical-card-grid: detects grid/flex parents with 3+ children sharing the
same structural fingerprint (icon + heading + paragraph template pattern).
- monotonous-spacing: regex on raw HTML collects padding/margin/gap values
(px, rem, Tailwind classes), rounds to nearest 4px, flags when >60% use
the same value with <=3 distinct values.
- everything-centered: regex counts text-align:center and Tailwind text-center
on text elements, flags when >70% of 5+ text elements are centered.
Also narrowed pure-black-white to only flag #000 as background color —
text-black, text-white, bg-white, and #fff are no longer flagged (too
common, per user feedback).
Extensive should-pass fixture covers: shadcn card sub-components, cards
with form inputs/dropdowns/code blocks/badges/accordions/tabs/images,
pricing cards, varied spacing, mixed centered/left-aligned layouts.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Element-level (jsdom) and regex (--fast) detection for:
- bg-black, bg-white, text-black: pure black/white
- text-white without dark bg class: pure white on light
- text-gray-*/slate-*/zinc-* on bg-{color}-*: gray on colored bg
- text-purple-*/violet-*/indigo-* on headings/large text: AI palette
- from-purple-* to-indigo-*: purple gradient
- bg-clip-text + bg-gradient-to-*: gradient text (already existed)
text-white is NOT flagged when paired with a dark bg class (bg-black,
bg-gray-700+, bg-blue-500+, etc.) since that's intentional contrast.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Five new detections:
- pure-black-white: flags #000/#fff in styles via regex (jsdom bg
resolution unreliable for this)
- gray-on-color: gray text (low chroma, mid luminance) on colored
backgrounds via getComputedStyle + ancestor bg walk
- low-contrast: WCAG AA violation (4.5:1 body, 3:1 large text) via
computed contrast ratio with resolved effective background
- gradient-text: background-clip:text + gradient combo via regex
(jsdom doesn't compute background-clip)
- ai-color-palette: conservative purple/violet accent detection via
regex on known hex values in prominent contexts
Background resolution handles jsdom limitation where background
shorthand isn't decomposed — falls back to parsing raw style attribute
for hex colors.
Color fixtures added for both should-flag (all 5 types) and
should-pass (tinted neutrals, good contrast, non-purple accents).
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Page-level typography checks (flat hierarchy, single font, overused font)
now only run on files that look like full pages (have <!DOCTYPE, <html>,
or <head> tags). Partials and components still get element-level border
checks.
isFullPage() strips HTML comments before checking to avoid false matches
on prose that mentions tag names.
Added partial-component.html fixture that has Inter, flat sizes, and a
side-tab border — verifies only the border is flagged.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Architecture simplified to two paths:
- HTML files: jsdom with getComputedStyle (resolves linked CSS, cascade)
- Non-HTML files: regex fallback (CSS, JSX, TSX, etc.)
- URLs: Puppeteer (unchanged)
- --fast flag forces regex-only for all files
Removed --deep flag (jsdom is now the default). Removed static mode
from browser script (always uses getComputedStyle — it's in a real
browser). Anti-pattern definitions split into:
- checkElementBorders() — shared element-level computed style checker
- checkPageTypography() — shared page-level checker
- REGEX_MATCHERS/REGEX_ANALYZERS — regex fallback for non-HTML
Browser script simplified from 470 lines to 250. CLI script reduced
from 810 lines to 440. Detection logic is now single-source for
jsdom/puppeteer/browser.
Fixtures now served via /fixtures/* route in dev server for proper
CORS handling of linked stylesheets.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Three new detections:
- overused-font: flags Inter, Roboto, Open Sans, Lato, Montserrat, Arial
as primary font-family or via Google Fonts imports
- single-font: file-level analyzer flags pages using only one non-generic
font family (needs pairing for typographic hierarchy)
- flat-type-hierarchy: file-level analyzer collects all font-size values
(px, rem, Tailwind text-* classes, clamp min/max) and flags when the
max/min ratio is below 2.0
Detection engine extended to support file-level analyzers alongside
line-level matchers. Typography fixtures added for both should-flag
and should-pass cases.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Anti-pattern detector script (source/skills/critique/scripts/detect-antipatterns.mjs):
CLI tool that scans files/dirs for UI anti-patterns via regex. Detects side-tab
accent borders and border-accent-on-rounded patterns across Tailwind, CSS, JSX.
Context-aware: skips safe elements (blockquotes, nav, inputs, code), neutral
colors, and adjusts thresholds based on border-radius co-occurrence.
- Browser visualizer (public/js/detect-antipatterns-browser.js):
Drop-in script that highlights anti-patterns directly in the browser with
labeled overlays. Two modes: "static" (regex, matches CLI) and "computed"
(getComputedStyle, catches CSS cascade). Scans both inline styles and
<style> blocks.
- Gallery of Shame (public/gallery.html):
Standalone page showcasing 11 AI anti-pattern examples with thumbnails
and links. Anti-pattern example pages updated from 1080x1080 Twitter
format to responsive layouts, labels removed, screenshots retaken at 16:10.
- Critique skill updated to run detector before manual review.
- Build system: skills now support scripts/ directories alongside reference/.
All 8 provider transformers refactored to use shared.js (DRY).
- 58 new tests covering detection logic, fixtures, CLI integration.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The typeset skill universally recommended fluid sizing via clamp(),
but no major app design system (Material, Polaris, Primer, Carbon)
uses fluid type in product UI. Now recommends fixed rem scales for
app UIs and reserves fluid type for marketing/content page headings.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Starts the response with a distinctive speed-line banner
to set the tone for the skill.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Two new mandatory sections in the skill:
1. Propose Before Building — must present 2-3 directions with
trade-offs and get user confirmation before writing code. This
skill has the highest misfire potential.
2. Iterate with Browser Automation — must use browser tools to
visually verify effects and iterate. Complex effects never work
on the first try.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Skill changes:
- Remove real-time collaboration section (product decision, not UI)
- Remove service workers / offline-first (product decision)
- Remove SharedArrayBuffer (too niche)
- Replace "application architecture" section with "performance-critical
UI" — focused on making existing features feel fast
- Add explicit note: this skill enhances how UI FEELS, not what a
product DOES
Beta label:
- Add betaCommands list to data.js
- Show BETA badge on /overdrive in command palette
- Show β symbol on periodic table element
- Show BETA badge on cheatsheet page
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The original was an effects-only skill biased toward creative
portfolios and visual spectacle. The rewrite addresses:
1. Scope: now covers ALL forms of technical ambition — performance
(virtual scrolling, WASM, Web Workers), interaction patterns
(View Transitions on dialogs, spring physics), real-time collab
(WebSockets, SharedArrayBuffer), and data visualization — not
just shaders and particles.
2. Browser support: removed Houdini Paint API (Chromium-only),
removed navigator.vibrate() (Safari never shipped, Firefox
dropped), added support notes for each technology tier.
3. Bias corrections: removed creative-portfolio-only framing,
removed "pick ONE hero moment" dogma, added examples for
functional UI (tables, forms, dialogs) and app architecture.
4. Structure: organized toolkit by what you're trying to achieve
(make transitions cinematic, push performance boundaries) not
by technology name.
5. Demo: replaced purple gradient AI slop with a structural
before/after showing view-transition-name on project cards.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
New skill that pushes interfaces past conventional limits with
bleeding-edge browser APIs: WebGPU shaders, scroll-driven animations,
View Transitions, generative art, spring physics, and more.
Key design decisions:
- Strong "when to use / when not to" guardrails
- Progressive enhancement is non-negotiable
- "Pick ONE hero moment" philosophy — restraint in choosing where
- The extraordinary/gimmicky line defined explicitly
- prefers-reduced-motion respect required
Also updates all counts to 20 commands across website, docs, and
plugin metadata.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Replace prescriptive 4pt px grid with flexible guidance (framework
scales, rem tokens, custom systems — consistency matters, not values)
- Add Flex vs Grid guidance — don't default to Grid when Flex is simpler
- Soften hierarchy advice: fewest dimensions needed, not "combine all"
- Remove margin-left optical trick (confusing, too niche)
- Remove touch target advice (belongs in /adapt and /audit)
- Add qualifier to icon centering (only adjust if confident)
- Fix position hierarchy to be about reading flow awareness, not rules
- Change hero metric from NEVER to nuanced DON'T (legitimate for real
data, problematic as default template)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Two new skills that expose frontend-design's typography and spatial
design references as actionable, diagnostic workflows:
- /typeset: assess and improve font choices, hierarchy, sizing, weight
consistency, and readability (closes#26)
- /arrange: assess and improve layout, spacing, visual rhythm, and
grid structure
Both follow the established skill pattern (assess → plan → execute →
verify) with MANDATORY PREPARATION and context gathering.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
The AI was inferring context from the codebase instead of running
teach-impeccable on cold start. Two changes:
1. Protocol in frontend-design now explicitly says "you cannot infer
context by reading the codebase" and marks teach-impeccable as
REQUIRED, not just suggested
2. Every skill's MANDATORY PREPARATION now reinforces: "if no design
context exists yet, you MUST run teach-impeccable first"
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Normalize already has its own "Discover the design system" step in
its Plan section that handles this. No need to duplicate it in the
preparation block.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
"What's essential vs nice-to-have" is product prioritization, not
design context. The protocol already covers audience and use cases.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Moves the protocol from a separate reference file directly into
frontend-design's main SKILL.md to avoid two-level indirection
(skill → frontend-design → reference file) that could cause the
AI to miss it. Also fixes leftover "shipping timeline" in the
built polish output.
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
Centralizes the duplicated ~15-line MANDATORY PREPARATION blocks from
10 skills into a single shared reference file in frontend-design. Since
all skills already depend on frontend-design, the protocol is
automatically available without build system changes.
Key changes:
- New reference: frontend-design/reference/context-gathering.md with
3-tier gathering: check instructions → check .impeccable.md → auto-
invoke teach-impeccable
- teach-impeccable now writes to .impeccable.md (provider-agnostic),
optionally also to the provider config file
- 8 existing MANDATORY PREPARATION blocks replaced with 1-line pointers
- 4 skills that lacked context gathering now have it (adapt, clarify,
polish, normalize)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
critique:
- Add MANDATORY PREPARATION block with context gathering and
{{ask_instruction}} — it was the only skill with a MANDATORY
PREPARATION section in other skills that lacked one
- Enforce frontend-design skill loading with "Do NOT proceed until..."
onboard:
- Add MANDATORY PREPARATION block — same gap as critique, onboard
jumped straight to assessment without gathering context first
- Add frontend-design skill loading
delight:
- Replace cliched loading messages ("Herding pixels", "Teaching robots
to dance", "Consulting the magic 8-ball") with product-specific
examples — these are exactly the kind of AI-generated copy the
suite's own anti-slop philosophy warns against
- Add explicit WARNING against these cliched patterns
- Change "Checkboxes that bounce" to "scale pulse" — bounce easing
contradicts the suite's consistent stance against bounce/elastic
curves (documented in motion-design reference and every skill
that mentions easing)
Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
- Filter teach-impeccable from {{available_commands}} (setup skill, not a fix)
- Soften wording: "prefer" our commands but also allow other installed skills
the LLM is sure exist, preventing false negatives while still preventing
hallucinated commands
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Adds a dynamic {{available_commands}} placeholder that the build system
replaces with the actual list of user-invokable skill names. Audit and
critique now explicitly instruct the LLM to only suggest from this list,
preventing hallucinated commands like /redesign.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
All commands are now skills with user-invokable: true. Source lives in
source/skills/{name}/SKILL.md. Added VS Code Copilot (.agents/skills/)
and Google Antigravity (.agent/skills/) transformers. All 6 providers
output to skills directories only — no more commands/prompts dirs.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Claude Code now ships a built-in /simplify command for code
simplification, which conflicts with our design simplification
command. Renamed to /distill across all source files, demos,
metadata, and documentation.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Remove the MANDATORY: Context Gathering section since it never
works when the skill is invoked from commands. Context gathering
is now handled by the commands themselves (bolder, quieter,
colorize, animate, delight, simplify).
The skill is now leaner and focused on design principles and
anti-patterns, matching the original Anthropic skill structure.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Commands now require context gathering before proceeding:
- bolder: target audience, use-cases, brand personality
- quieter: purpose, audience, what's working
- colorize: brand colors, audience, domain appropriateness
- animate: personality, performance budget, audience
- delight: brand personality, emotional context, appropriateness
- simplify: what's essential vs nice-to-have
Each command now:
1. Gathers context from thread/codebase first
2. MUST ask user if inference confidence is medium or lower
3. MUST use frontend-design skill before proceeding
4. Has visual separator (---) as hard gate before design work
This prevents AI slop by ensuring commands have proper context
before making design decisions.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add "MANDATORY: Context Gathering (Do This First)" section with 3 procedural steps
- Require scanning codebase for README, brand guidelines, design tokens before design
- Must answer: target audience, brand personality, existing tokens/colors/fonts
- Hard gate: STOP and ask user if ANY question is unclear
- Visual separator (---) prevents proceeding without completing steps
- Preserves full Tone options and Design Direction content
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Monospace is fine when used intentionally - the issue is lazy "dev tool"
defaults, not the typeface itself.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Added explicit AI SLOP TRAP warning before planning
- Removed "monospace as accent" suggestion (contradicts skill)
- Replaced glassmorphism with grain, halftone, duotone alternatives
- Added anti-slop caveats to shadows and borders suggestions
- Added "NOT AI slop" as first verification check
- Added "the test" to verify section
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Restored DO/DON'T format in individual sections (required for website)
- Added new "Visual Details" section for glassmorphism, sparklines, etc.
- Added new AI slop patterns: gradient text, dark mode with glowing accents,
hero metric layout, identical card grids, thick colored border on one side
- Added "The AI Slop Test" section for commands to reference
- Commands now reference "DON'T guidelines" (one source of truth)
- Build now shows 16 pattern categories (was 0)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Removed scattered DON'Ts from individual Frontend Aesthetics sections
- Created ONE "Anti-Patterns (CRITICAL)" section with two sub-sections:
- AI Slop Tells: cyan-on-dark, gradient text, glassmorphism, hero metrics, etc.
- Design Anti-Patterns: gray on color, nested cards, bounce easing, etc.
- Added new pattern: "Rounded element with thick colored border on one side"
- Commands now reference the skill's Anti-Patterns section instead of duplicating
- Easier to maintain: one place to update
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Critical addition: Both commands now check for AI-generated aesthetic tells:
- Cyan-on-dark, purple-to-blue gradients, neon accents
- Gradient text, glassmorphism, hero metric layouts
- Monospace typography, sparklines as decoration
- Identical card grids, generic fonts, rounded rectangles
Reports now start with "AI Slop Verdict" - pass/fail on distinctiveness.
This is the most important quality signal for shipping work that doesn't
look like every other AI-generated interface.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Commands now say "Use the frontend-design skill" instead of passive note
- Skill's Design Thinking section now has explicit MUST-know checklist
- Added ask instruction to bolder command for context gathering
- AskUserQuestionTool instruction is now more prominent and imperative
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The /review command conflicts with a built-in Claude Code command.
Renamed to /critique which better reflects design terminology.
- Rename source/commands/review.md → critique.md
- Update command content and descriptions
- Rename demo file and update references
- Update data.js command mappings
- Update case study example in index.html
- Rebuild dist files for all providers
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Rewrite README to be user-focused (move dev info to DEVELOP.md)
- Credit Anthropic's original frontend-design skill
- Change license from MIT to Apache 2.0 for consistency
- Add NOTICE.md with proper attribution
🤖 Generated with [Claude Code](https://claude.com/claude-code)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
