Skip to content

CHANGELOG.md

Latest commit

 

History

History
52 lines (41 loc) · 3.69 KB

CHANGELOG.md

File metadata and controls

52 lines (41 loc) · 3.69 KB

Changelog

[1.0.0] — 2026-04-18

All 52 skills complete

CANON ships with every planned skill written, every harness bundle compiled, and a working detector CLI.

Phase 3 — themes, content, accessibility deep-dive (11 skills):

  • canon-dark-mode — lightness inversion (not hue shift), surface stacking, semantic tokens, contrast re-audit, three-state toggle
  • canon-density — compact/default/comfortable modes, spacing multiplier, font size unchanged, mobile floor preserved
  • canon-rtl — CSS logical properties, mirroring rules (what flips, what doesn't), bidi text, <bdi> element
  • canon-i18n — string externalization, ICU MessageFormat pluralization, Intl API formatting, 40% text expansion budget, no concatenation
  • canon-brand-voice — voice attributes with do/don't pairs, tone spectrum per context, banned phrase list, AI copy guardrails
  • canon-content-strategy — content-first design, progressive disclosure, scannability, content inventory, heading-as-statement
  • canon-information-architecture — flat vs deep hierarchy, URL design, labeling in user language, card sorting/tree testing validation
  • canon-keyboard — tab order = visual order, roving tabindex for composites, shortcut conventions, no tabindex > 0
  • canon-screen-reader — semantic HTML first, heading hierarchy, ARIA only when HTML fails, live regions, .sr-only pattern
  • canon-focus — :focus-visible spec, 2–3px ring at 3:1 contrast, double-ring technique, SPA focus management
  • canon-color-blindness — color never the only signal, CVD-safe palettes, grayscale squint test, chart shape/pattern rules

Phase 4 — patterns and engineering (13 skills):

  • canon-micro-interactions — trigger-feedback loop, 100–300ms budget, reduced-motion fallback, what earns animation
  • canon-onboarding — self-evident UI > smart defaults > empty states > coachmarks > checklists > tours > video
  • canon-error-handling — what/why/how anatomy, inline vs toast vs banner vs modal, retry with backoff, error boundaries, offline banner
  • canon-illustrations — when illustrations help vs clutter, one style per product, sizing by placement, decorative vs informative alt
  • canon-cursors — cursor-affordance pairing table, hover+cursor rule, custom cursor rules, no pointer on non-interactive
  • canon-grids — Grid for 2D / Flexbox for 1D, auto-fill vs auto-fit, gap from spatial scale, subgrid for alignment
  • canon-data-viz — big number + context pattern, sparkline spec, progress bar ARIA, metric comparison, dashboard composition
  • canon-bundle-size — 200KB JS budget, heavy dependency alternatives, code splitting, lazy loading, Brotli
  • canon-images — AVIF/WebP format tree, srcset + sizes, LCP image treatment, width/height for CLS, lazy below fold
  • canon-fonts-loading — font-display: swap, preload critical weight, WOFF2 only, subsetting, FOUT reduction with metric overrides
  • canon-seo — semantic HTML is SEO, essential meta tags, JSON-LD structured data, sitemap, CWV as ranking factor
  • canon-meta-og — OG tags, Twitter cards, 1200×630 image spec, dynamic OG generation, validation tools
  • canon-favicons — minimal set (ico, svg, apple-touch, manifest), SVG dark mode, design rules at 16px, no transparency on iOS

Totals at v1.0:

  • 52 skills (12 fundamentals + 16 components + 4 themes + 3 content + 4 accessibility + 7 patterns + 6 engineering)
  • 19 commands
  • 89 detector rules
  • 15 harness bundles
  • 32 tests passing

[0.2.0] — 2026-04-16

Phase 2: 16 component-tier skills, detector expanded from 50 to 89 rules. See previous changelog for details.

[0.1.0] — 2026-04-16

Initial release: 12 fundamental skills, 19 commands, 50 detector rules, 15 harnesses, 32 tests.