778 Commits (6b54f47ebec9a651d1f0b4d236a0fb965e019517)

Author	SHA1	Message	Date
dev	1186d794cd	refactor(words): close audit debt D1 (honest cast) + D2 (dedup image picker) ... - D1 (soma) block-tree.ts:104 — replace `doc as never` (which disabled all type checking on the descend container) with `doc as unknown as {...}`, the honest cast that still asserts the permissive descend-node shape childStacks() reads. - D2 (eidos) — extract the duplicated ~25-line "file picker → FileReader → data URL" flow (copy-pasted in words-block-gutter.svelte and words-block-panel.svelte) into a shared `pickImageFile(doc, onPick)` helper in words-image-file.ts. Both consumers now call it; zero inline FileReader blocks remain. Kept as a plain helper (not ActiveDom) — a transient <input type=file> + FileReader is not instrumented by active-dom, so a direct Document touch is correct; the caller resolves the right Document. Gates: npm run check 1 error (pre-existing grafito, not Words) · words soma 452/452 · prettier clean. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>	4 months ago
dev	49802e9257	fix(words): close 4 cross-layer audit findings (morfo/soma/eidos) ... Full audit of the Words component (3 parallel explore agents + deterministic gates) surfaced 4 carried-over defects, all verified and fixed: - R2 (soma) engine/operations/visual.ts:147 — the `node as Record` cast in updateBlockAtPath failed because WordsNode includes WordsDocument (no index signature). Use the double cast `as unknown as Record`, consistent with the sibling op at line 118. Clears the only Words type error in `npm run check`. - R1 (morfo) components/words.ts — the `data-words-node` enum was conflated with `data-words-block`: it listed block TYPES (paragraph/heading/quote/ code/image) that nothing emits on that attr, and OMITTED the real structural kinds the engine writes (`block`/`inline`). Aligned to the engine's WordsDomNodeKind. eidos-lint invalid selectors 2 → 0; verified in the browser that top-level blocks carry node='block' and the active-block outline still paints. - R3 (eidos) langs-inspector.ts — `LABEL_COLUMN_BLOCKS` had no bundle entry, so the column "Blocks" panel title fell back to English in ES. Added label.column-blocks + label.add-block. Verified: panel now reads "Bloques". - D3 (morfo) components/words.ts — declared `expression: 'family-default'` to make the no-sema-pack choice explicit; `morfo:vocabulary` no longer warns about the missing pack/expression. Gates: npm run check 1 error (pre-existing, web/routes/temas/grafito — not Words) · words soma+morfo 514/514 · eidos-lint words invalid 0 · prettier clean · morfo:vocabulary exit 0. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>	4 months ago
dev	8090efc833	docs(theming): align THEMING.md + README with current color/scaling engine ... Reconcile the theming reference with the engine as already committed (color-model redesign, scaling axis, P2 fixes) so the doc is internally consistent: - TOC: add §23 (scaling), §24 (P2 corrections), §25 (color model); fix the §21 entry to its resolved heading. - §4: intent->scale mapping was stale (risk->amber not orange, loss->plum not purple); note intents auto-derive via CANONICAL_INTENT_SCALES (identity = step 9, cross-ref §25); correct token counts (31-scale palette = 744 --scale-* tokens; 9 roles = 216 primitives). - §3 / anti-pattern G: "30 escalas" -> 31; reconcile "never add a scale" with §25.7 (a brand theme brings its own palette). - §21 / §22: mark the two-level color RFC as resolved in §25 (anchor model discarded); drop the stale present-tense "vigente" claim. - README reference table: add §23 / §24 / §25 rows. Docs only -- no engine or demo changes. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	4 months ago
dev	152d2ad960	Theming: Radix-parity palette + intent auto-derivation + color-model docs ... Two-level color model settled (THEMING.md section 25), replacing the anchor RFC: the palette is the source (scales, directly usable, designable); hierarchy roles alias scales explicitly; intents auto-derive from the palette by the book's canonical convention. - Palette library expanded 12 -> 31 scales at Radix Colors parity (exact values): radix-scales.ts (19 added: mauve/sage/olive/sand/tomato/ruby/crimson/plum/ violet/iris/indigo/jade/grass/brown/sky/mint/lime/gold/bronze) spread into base.ts. Each directly usable as --scale-{name}-{step}. - Intent auto-derivation: CANONICAL_INTENT_SCALES (neutral->gray, affirm->teal, fulfill->green, risk->amber, threat->red, loss->plum) + completeColorRoleMap. Intents omitted from a theme role map fill from the convention (identity = step 9); slots derive normally; override optional. ColorRoleMap: hierarchy required, intents optional. - Validation: hierarchy roles required; omitted intents validate the canonical scale exists in the palette. - index: export ScalingKey / SCALING_KEYS. - docs: THEMING.md section 25 (full color model + decisions), section 21 marked resolved, COLOR_MODEL_RFC resolved (anchor rejected). - test: base library asserts 31 scales x 12 steps. Verified: npm run check (0 new errors), vitest eidos (0 new regressions), browser (intents auto-derive: affirm=teal #0E9384, risk=amber #DC6803, loss=plum #7A3AAD at step 9). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	4 months ago
dev	66ce4f354d	Theming: on-solid contrast by luminance (P2-2) + translucent role surfaces (P2-4) ... Two audit P2 quality defects, fixed at engine level. P2-2 - on-solid text illegible on light solids: The `contrast` slot defaulted to `--color-content-on-solid` (white) for every role. On light solids (amber/yellow, risk=orange ~2.3:1) white is sub-AA. The engine now picks by WCAG contrast (gamma-linearized) of the role step-9: when white fails (<3:1) it uses `--color-content-on-solid-contrast` (a dark, new OPTIONAL `content.onSolidContrast` semantic, #1c1917 in base). Only risk flips to dark in base (5.89:1); purple/red/teal/green keep white (convention, >=3:1). An explicit `slots.contrast` override is still honored verbatim. P2-4 - opaque tinted soft surfaces: The soft variant tint (Button + Badge `{role}-soft-bg`) was opaque (step-1 track + opaque color-mix hover) so it did not composite over non-uniform backgrounds. New derived tokens `--color-{role}-surface` (= a2) + `--color-{role}-surface-hover` (= a3) are translucent by construction (compositing-inverse alpha). Button/Badge soft consume them. Toast/Tabs untouched - they are cards, opacity is correct. - config-types: optional onSolidContrast + CONTENT_COLOR_OPTIONAL_KEYS. - config: content keySet allows the optional key; validator value-checks it. - render-css: wcagContrastRatio/wcagRelativeLuminance; luminance pick; emit on-solid-contrast + surface/surface-hover. - contract: on-solid-contrast + surface tokens per role. - themes/base: onSolidContrast #1c1917 (light + dark). - recipes/base: Button + Badge soft-bg -> surface, soft-bg-hover -> surface-hover. - docs: THEMING.md section 24 + section 22 item 2; audit P2-2/P2-4 resolved. Verified: npm run check (0 new errors), vitest eidos (0 new regressions), browser runtime (risk 5.89:1 dark text, surfaces translucent rgba). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	4 months ago
dev	bceaef41e0	Theming: add `scaling` zoom axis (Radix parity), separate from density ... Introduce a global zoom axis independent of density, in parity with Radix Themes' `scaling` (90/95/100/105/110%). Scaling zooms px metrics INCLUDING typography (font-size, icon-size, space, control-height); density only moves layout rhythm + control height and leaves text fixed. The two axes compose multiplicatively. - config-types: SCALING_KEYS / ScalingKey / DEFAULT_SCALING; DensityPrimitiveSet drops the dead `scale` + `contentScale` (kept spaceScale, controlScale). - primitives/static: STATIC_SCALING (0.9..1.1). - render-css: appendScaledMetricDeclarations wraps metrics in calc(<raw>[ * var(--density-x-scale)] * var(--scaling)); appendScalingDeclarations emits --scaling-{key} + --scaling default; renderScalingBlocks emits [data-scaling] blocks. line-height/radius/border/shadow excluded. - config + contract: prune the removed density scalars. - active-eidos: `scaling` / `scalingSource` options, getScaling() on the preference source, data-scaling projection + dispose cleanup. - docs: THEMING.md section 23 + 20.1 reconcile; README density/scaling; SCALING_RFC.md. Verified: npm run check (0 new errors), vitest eidos (0 new regressions), browser cascade at 90/100/110 scales font/space/control x0.9/x1.1 and leaves radius/border fixed. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	4 months ago
dev	98d6799ad0	Theming: generate alpha scales via compositing inverse (P1-1) ... Replace the synthetic step-9-at-opacity alpha ramp with a real compositing inverse: each --scale-X-aN is the translucent color that, painted over the scale's background (white for light scales, black for dark, by step-1 luminance), reproduces the opaque solid step N (Radix model). Computed in render-css.ts for ANY theme (base + custom), so alpha is consistent-by-construction with the solid scale and cannot drift. alphaScales opt-in still wins verbatim; non-hex solids fall back to the legacy color-mix ramp. Verified: --scale-purple-a9 = rgb(92 0 173 / 0.6941) over white = #8e4ec6. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	4 months ago
dev	ef9f589d4c	docs(theming): clarify synthetic alpha-scale model + alphaScales opt-in (P1-1) ... Alpha steps (--scale-*-aN) are unconsumed today (only a test + README ref). The step-9-at-opacity synthesis is a valid 'accent-at-opacity' ramp default, not a bug; it is not a Radix-style reproduction of the solid scale. Documented the model, its limitation, and the already-wired alphaScales opt-in in appendColorAlphaScaleDeclarations. A build-time compositing-inverse generator is deferred until alpha is actually consumed. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	4 months ago
dev	c8fa6e87ed	docs(theming): fix §19 reference to nonexistent ActiveEidos.setOverrides() ... The runtime token-override API is setCssVariables() / clearCssVariables(); setOverrides() never existed. A dev copying §19 would hit a runtime error. (audit P2-5/A1) Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	4 months ago
dev	3adc93b2d1	Banner: route palette through recipe forwarders (P1-5) ... - Add 40 recipes.banner.{intent}-{slot} forwarders (= var(--color-{intent}-{slot})) - banner.css reads --banner-{intent}-{slot} instead of foundation --color-{intent}-* directly - Banner is now retintable per-component via recipes.banner.* like button/badge; keeps data-intent (no API change) - Regenerated base.css; recipe contract balanced (40 declared = 40 consumed), no visual regression Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	4 months ago
dev	218ccfcc5e	Theming: untitled-ui demo theme, engine fixes, audit + P0/P1 fixes ... - New untitled-ui demo theme + theming showcase route (web/routes/temas) - Engine: density tokens scale via calc(var(--density-*-scale)); contrast slot defaults to on-solid - Eidos theming audit (THEMING_AUDIT_2026-06-01.md) + two-tier color-model RFC (COLOR_MODEL_RFC.md) - P0: replace phantom foundation tokens in recipes + component CSS — shadows -> semantic scale (subtle/raised/overlay), add --space-7, --border-width-strong -> thick; neutral solid contrast -> step 12 (AA in light + dark) - P1: validate hex color values + per-theme semantic completeness; regression tests for contrast, density, dark output, and TSC multi-part/composition emission Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>	4 months ago
dev	9324764675	feat(words): inspector UX overhaul — id-based active block, controls, chrome ... Active-block addressing by materialized-path id (words-activate.svelte + engine/block-tree.ts): click sets the active block explicitly; inspector follows it, immune to selection churn. Breadcrumb navigates ancestry. Inspector controls: - ToggleGroups (font / weight / align / border) always keep one option selected; the default value stores as undefined to keep the model clean. - Border gains a `none` chip that clears the border (BorderChipValue). - Image full-width is a Switch; image src + column width/gap validate on commit (sanitizeWordsUrl, CSS-length/flex checks) with a [data-invalid] ring. - Corner-radius + border-width are slider-only (WordsNumRow `sliderOnly`): the paired numeric stepper was redundant. - Numeric fields clamp to their range (line-height ≤ 2.5, font-size 12-40). Inspector chrome: - Unified field frame + single accent focus ring (no inset outline over text, no ghost/black-border mix). - Section dividers + uppercase accordion headers; styled breadcrumb row. - Depth-tied accent (primary/affirm/loss/risk) shared by header mark and the canvas active-block ring. - Selection ring is an ::after pseudo-element with a fixed radius, so a block the user rounded no longer rounds its selection indicator. - Raw hex/rgba replaced by --color-*-solid tokens (theme-aware, portal-safe). Columns: empty on insert, gap control, per-column width; Add-block removed from the Columns BLOCK panel (lives in the column inspector). Shift+Enter soft break renders via white-space: pre-wrap. npm run check: 1 pre-existing baseline error (visual.ts:147). Words soma 452/452. Verified end-to-end in the browser. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>	4 months ago
dev	7be1d54f89	fix(words): inspector follows every click — drop sticky atomic-block selection ... The provider had a "smart" rule in syncSelectionFromDom that kept the atomic-block highlight (and therefore the Inspector's panel) sticky while the caret moved to a different block AS LONG AS that next block was inside the same TOP-LEVEL wrapper (the rationale: editing the trailing paragraph after an image-in-column shouldn't kick the user out of the image inspector). In practice users got the opposite of what they expected: they clicked on something new, the inspector didn't follow. They thought the editor was broken. Drop the smart rule. Every real selection change now also clears the atomic-block highlight — the inspector follows the caret, period. If the user wants the atomic's panel back they click the figure again (which `handleAtomicClick` re-selects). We still skip the clear when `sameWordsSelection` reports no-op so the synthetic selectionchange echo that fires right after `selectAtomicBlock` (which doesn't move the DOM caret) can't wipe the highlight a tick after being set. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	e81b53e806	refactor(words)!: column block insertion moves to inspector (React Bricks pattern) ... The floating in-canvas `+` overlay for column inserts was fundamentally at odds with contenteditable semantics — it needed pointer-events:none to let caret/text events pass through, but that ALSO disabled CSS :hover detection, and any visible variant covered the column's auto- trailing paragraph (the escape hatch the engine adds after atomic inserts), hijacking clicks intended for typing. Adopt React Bricks's pattern: container-block child insertion goes through the sidebar/inspector, not via floating overlays on the canvas. - New "Añadir bloque" dropdown per column in the Columns inspector panel (`words-block-panel.svelte`). Lists every insertable block type the engine knows (paragraph, headings, lists, image, divider, callout, etc.). Picking one dispatches `insertBlockInColumn`, appending it to the chosen column. - Inspector lives outside the contenteditable, so the dropdown doesn't fight focus traps, doesn't cover content, doesn't need pointer-events gymnastics. Zero of the bug classes we hit. - Deleted `words-column-inserter.svelte` and its mount in `words.svelte`. The `insertBlockInColumn` engine op stays — it's the right primitive, just driven from a different surface now. - Added `ARIA_ADD_BLOCK_TO_COLUMN_N` and `LABEL_ADD_BLOCK` to the inspector langs catalog. Top-level block insertion (between rows) keeps using the canonical gutter handle `⋮⋮ → Insert below` and the slash menu (for in-flow keyboard users). Inside-column editing keeps working as before: click on the block, type. Enter at end of a paragraph creates a new paragraph below within the column. Breaking: anyone who imported `words-column-inserter.svelte` directly is broken. Nobody outside this folder did. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	eb293b4e50	fix(words): column inserter overlay only covers bottom strip on non-empty columns ... The wrap was rect-matched to the whole column rect, which meant the `+` trigger (positioned at the bottom of the wrap, pointer-events:auto on hover) overlapped the column's TRAILING paragraph — the empty escape-hatch the engine auto-adds after atomic inserts (image, divider). Clicking on that paragraph to start typing hit the `+` trigger and opened the dropdown instead. Fix: for non-empty columns, position the wrap as a thin 28px strip at the column's bottom (where the `+` lives) instead of covering the whole column. The trailing paragraph above the strip is now clickable. Empty columns keep the full-column wrap (so the centered `+` invitation remains). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	8cdc4e3b9e	fix(words): column inserter `+` reveals on hover for non-empty columns ... `pointer-events: none` on the inserter wrap (required so caret/text events pass through to the contenteditable below) ALSO disabled CSS `:hover` detection — the browser doesn't dispatch `mouseover` to pointer-events:none elements, so the wrap's `:hover` styles never fired and the `+` trigger stayed at opacity 0 forever. Fix: track hover via JS — listen to `mouseenter`/`mouseleave` on the column elements directly, stash the key in `hoveredColumnKey`, expose to CSS via `data-hovered` on the wrap. CSS now uses `[data-words-column-inserter][data-hovered]` instead of `:hover`. Effect: hovering a non-empty column now reveals the `+` at the bottom of the column, allowing users to add more blocks below the existing content. Same fix applied to the empty-column reveal-on-hover (the opacity-1 boost when fully hovered). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	eb946148cd	fix(words): column inserter — caret lands in new block, typing replaces stub ... Three composing bugs were blocking the `+`→pick-block→type flow: 1. **Path 5-segmentos en `insertBlockInColumn`** — the op passed the path TO the text node (`[colsIdx, colIdx, innerIdx, 0]`) to `pointFromInlineTextOffset`, which expects the path to the CONTAINER block. The helper then appended its own inline-index, producing a stale 5-segment path that no `[data-words-path]` selector matches. Fix: pass `[colsIdx, colIdx, innerIdx]` for heading/paragraph/quote/code/callout, `[colsIdx, colIdx, innerIdx, 0]` for list (the first list-item IS the container). 2. **Selectionchange echo overwriting the model** — `api.focus()` landed the caret at doc-start (browser default for focusing a contenteditable without a prior selection), fired selectionchange, `syncSelectionFromDom` overwrote the model's intended post-insert selection with `0.0` before `restoreDomSelection` could run. Fix: capture `intended = api.selection` immediately after the op, then re-apply via `setTimeout(0)` AFTER the echo handler has settled. 3. **DropdownMenu focus trap intercepted `api.focus()`** — the FocusScope's capture-phase `focusin` bouncer kept yanking focus back inside the menu. Fix: `trapFocus={false}` on Content. Also expose `api.restoreCaret()` (provider line 1274) for any consumer that needs to force-write the model selection to DOM after awaiting enough ticks for new blocks to mount. Verified end-to-end in Chrome: empty column → click `+` → pick Heading → "Title" stub selected → typing immediately replaces with user's text. Inspector switches to Heading panel. Doc title and rest of the doc untouched. CONTINUE.md P0 — RESUELTO. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	bcad6647a7	feat(words): column inserter + provider DOM-selection sync after commands ... Three architectural pieces for inserting blocks into a column slot from an overlay button, plus the missing DOM-selection sync that any imperative consumer of `applyCommand` needs. - Engine op `insertBlockInColumn` returns `{document, selection, activeMarks}` in one transaction (Tiptap-style). Atomic blocks land with a trailing paragraph escape hatch + caret there; text-bearing blocks select any stub text ("Title", "List item") so the next keystroke replaces it Notion-style. - Provider `applyCommandWithOptions` now schedules `restoreDomSelection` via tick when the command changes the model selection (typing-batch excluded — the browser already placed the caret). Was the hidden gap: overlay buttons, drag-drop, slash menu, the new column inserter, all updated the model but the DOM caret stayed wherever the user last clicked, breaking subsequent text editing. - `words-column-inserter.svelte` rebuilt around a busy guard with a hard 250ms safety timeout (the previous pendingInsert + onCloseAuto Focus pattern could leave the `+` button dead forever if the dropdown's teardown swallowed the close callback). Plus type sync: `WordsProviderSnippetProps` now declares `selectedBlockPath`, the second arg of `selectAtomicBlock`, and `setSelection` — they were exposed by the runtime but missing from the type, breaking typecheck on eidos consumers. Demo carries a `columns` block in the initial doc as a permanent test fixture for column-related fixes. **Known issue documented in CONTINUE.md P0:** typing inside a `columns` block does NOT insert — selection sync (`syncSelectionFromDom`) isn't mapping nested paths (`12.0.0.0`) to the model correctly. The inserter flow is wired correctly; once the path encoding for nested selections lands, the full Notion-style "click + → pick Heading → type" flow works end to end. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	da8d03085b	feat(words): inspector ColorPicker integration + 4 root-cause picker fixes ... Lands the eidos `<WordsColorRow>` (Text / Background pickers inside the block inspector) on top of the new base-Block words engine + drops 3 obsolete audit MDs. ColorPicker fixes surfaced while wiring it into the inspector — all documented in `soma/components/color-picker/README.md` §Integration pitfalls: 1. Eidos wrapper now declares `format = $bindable('hex')` (matches soma's default). Without it, `bind:format={undefined}` threw `props_invalid_value` on every mount → render loop. 2. Eidos wrapper now forwards `ref` to `ColorPickerProvider.create`. Without it, `attachRef` was never built and any `runtime.trigger` targeting `provider` threw `SomaRuntimeTargetError`. 3. `triggerClose` falls back to the picker provider's own DOM when `runtime.partRef('content')` returns null (content is registered on the Popover's runtime, not the picker's — re-exported part). 4. WordsColorRow draft-pattern: `$effect` reads draft inside `untrack` so it doesn't overwrite mid-drag value, and `onValueChange` catches the `Clear` programmatic transition (`onValueChangeEnd` doesn't fire on Clear). CSS: `contain: layout style` on `[data-color-picker-area]` and `scrollbar-gutter: stable` on the picker popover content — stops the area from shaking during drag when the trigger's ValueText hex changes width or the popover scrollbar toggles. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	73636d07a9	refactor(words)!: F1 — rework engine to base-Block model, drop v2 + sema ... Per the user-directed rework. The engine is rebuilt around a single base `Block` carrying the common properties (id, align, margin, padding, background, text color, border), with each specialized block (paragraph/heading/quote/code/list/table/image/divider/callout — plus list-item/table-row/table-cell) EXTENDING it. The old `visual` sidecar + per-type `Pick<>` whitelist + loose `textAlign` are gone; common props are first-class and universal. Naming: dropped the `v2` noise everywhere — `operations-v2/`→ `operations/`, `types-v2`/`render-v2`/`validate-v2`/`serialize-*-v2`→ clean names, `*V2` symbols→bare, `WordsEvalIntent`→`WordsIntent`. Sema: removed entirely from the editor — deleted `sema/components/ words.ts` + `engine/sema-parity`, dropped the registration from the sema barrel + the demo layout, no `runtime.trigger`/semantic events. Visual layer: deleted `eidos/components/words/*` (the bespoke chrome) and stubbed the demo — both rebuilt next on the new model (F2 render by components, F3 gutters/bubble/drawer, all eidos). Engine tests 400/400 green; `npm run check` 0 errors (5142 files). Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	4 months ago
dev	1bb16ff636	feat(words): border-light chrome restyle (declutter) ... The editor read as "poor" largely from border clutter — hairlines everywhere. Strip incidental borders and let tonal surfaces + spacing carry the structure: - Toolbar: drop the bottom hairline (its surface tone separates it). - Drawer: drop the seam line (both sides) — tonal separation only. - Drawer panels: drop inter-panel divider lines (scope accent + spacing). - Code block: drop the outline — the overlay surface fill defines it. - Table: no outer box, no vertical lines — light horizontal row rules only + airier cells. - Drawer insert items: borderless soft tonal tiles (hover deepens with the accent) instead of bordered boxes. Kept the intentional lines (quote accent bar, the silver margin rule). Verified in browser: no console errors; table is horizontal-rule-only, insert tiles borderless. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	4 months ago
dev	8afe2728bb	feat(words): comfortable reading scale for the content surface ... The editor used UI-cramped metrics (14px / 1.25 line-height / 8px block gap) which read as a form field, not a document. Give the CONTENT (not the chrome) a reading scale: body 14→16px, line-height 1.25→1.6, block gap space-2→space-3 (headings derive 1.5x for top margin). Chrome keeps its smaller --_words-font-size. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	4 months ago
dev	9e13d0083e	fix(words): refine the left margin band per feedback ... - Decouple band width from text inset: `--_words-rail-width` (1.25rem, the thin band) + `--_words-rail-gap` (1.25rem, breathing room) → text is no longer glued to the band edge. - Narrow the band: 2.5rem → 1.25rem (20px). - Lighter tones: rail-bg #ededed → #f5f5f5, rail-border #cfcfcf → #e3e3e3. Eidos CSS regenerated. Verified: band 20px + 1px #e3e3e3 edge, then 20px air before text at 40px. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	4 months ago
dev	540c35e679	feat(words): gutter rework — drop "+", move insert into drag menu, silver band ... User-directed editor rework, step 1 (left gutter): - Remove the "+" inserter button from the block-handle gutter. Its action moves into the drag-handle menu as a new "Insert below" item (folded into `act('insertBelow')`); the standalone `insertBelow` helper is gone. The separate seam inserter (words-block-inserter) is untouched. - Turn the left gutter into a soft silver margin band: narrow it (`--_words-gutter-size` 3.5rem→2.5rem) and paint it via a background-gradient on [data-words-content] — `--words-rail-bg` #ededed band with a 1px, slightly stronger `--words-rail-border` #cfcfcf right edge. Background only, so the contenteditable text / selection is undisturbed. Recipe tokens softened + eidos CSS regenerated. Verified in browser: gutter shows only the grip on hover (no "+"); the menu reads Insert below / Move up / Move down / Duplicate / Delete and "Insert below" inserts a paragraph (7→8 blocks) and closes the menu; the silver band renders at 40px with the 1px border. npm run check: 0 errors. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	4 months ago
dev	005239c732	feat(words): inline heading-level control in the Block drawer panel ... Adds an H1/H2/H3 toggle to the drawer's Block panel, shown only when the caret is in a heading (currentBlock === 'heading'). The toggle reflects currentHeadingLevel and dispatches the structured `setBlock` command with the numeric level via applyCommand: applyCommand({ type: 'setBlock', block: 'heading', level }) runCommand / WordsCommandName has no per-level id (heading-1/2/3 are slash/insert-menu ids), so the structured command carries the level; `as const` keeps each level a 1|2|3 literal. Verified in browser: caret in a heading shows "Heading level"; the toggle reflects the active level and converts bidirectionally (H1→H3→H2) with the toggle updating. npm run check: 0 errors. Also closes the stale "drag handle outside the rail over <code>" P1: it does not reproduce — the rail was removed in the gutter redesign (5450d5f7) and the handle now anchors to contentRect.left+8 for every block type (verified: identical left over code and paragraph). Both P1 items marked resolved in continue.md. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	4 months ago
dev	cc4412b8e9	feat(words): complete image visual panel (R5 / POLISH-1b) ... The drawer image panel already had a corner-radius slider and shadow/border toggles. Adds the controls that were still missing so the V2 `block.visual` sidecar is fully editable from the UI: - shadow blur slider (shown when shadow is on) → visual.shadow.blur - shadow color picker → visual.shadow.color - border color picker (shown when border is on) → visual.borderColor - background color picker (always, with clear) → visual.background All route through a small reusable `imageColorField` snippet (preset swatches + native <input type=color> + optional clear), mirroring the format panel's text-color UI, and dispatch `setBlockVisual`. The engine already merges + renders these via visualToStyle onto the <figure>. Verified in browser: panel renders all four new controls when an image is selected; clicking a Background swatch updates the figure (background-color:#22c55e) while preserving radius/border/shadow (shallow-merge intact). npm run check: 0 errors (26 pre-existing warnings, none new). Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	4 months ago
dev	4a0661a53b	chore(words): sweep D2 orphans from eidos CSS + README ... Removes now-dead selectors left after V2 dropped table striped/compact and cell tone/header (D2). Nothing emits these attrs anymore: - `[data-command='table-cell-tone-*']` toolbar rules - `table[data-words-table-striped]` / `[data-words-table-compact]` rules - `[data-words-cell-tone='muted'|'accent']` cell + focus-within rules - the dead `[data-words-cell-header]` alternates (headers render as <th>); `tr:hover td:not([data-words-cell-header])` simplified to `td` README updated: tables list only the live formatting (header row/column, cell align, vertical align); HTML/Markdown documented as export-only, import as JSON + plain text. ARCHITECTURE_PROPOSAL keeps its mentions — they document the decision to drop these, not stale features. Verified in browser: sample 2x2 table renders, <th> keeps the accent tint + weight via :where(th), cell align/vertical intact, no console errors. CSS + Markdown only — no type surface touched. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	4 months ago
dev	10ed029b91	refactor(words)!: remove V1 engine — V2 is now the sole engine ... Deletes the V1 rich-text engine and the extensions/ tree. The public document model flips from WordsDocument to WordsDocumentV2 (D1: V2 puro, sin legacy — no backward-compat importer, V1-persisted docs no longer load). Per D2, table striped/compact/cell-tone are dropped entirely (props, commands, morfo attrs). - Delete engine/{document,operations,commands,history,input,render, find-replace,serialize-{html,json,markdown,text},dom-selection,index} and the whole extensions/ tree (registry + image + table). - Complete the V2 ops under operations-v2/: port commands, history, input, find-replace, render-dom, serialize-text, extra-ops, insert-block-types, markdown shortcuts, factories, dom-selection; land R4A.2-R4A.8 (text/delete/marks-ops/paragraph-ops/delete-ops/ block-format/list-ops/link-ops/table-ops/selection-walkers). - Rewire provider, exports, types, morfo/words.ts, eidos word components, soma words.svelte, the content-test harness, and the /uix/components/words demo to V2. - Import accepts JSON (JSON.parse + validateWordsDocument gate) and plain text only; HTML and Markdown remain export-only (lossy, no faithful inverse parser). Verify: npm run check 0 errors; words scope 455/456 (lone failure is the sema-parity 5s timeout flake under parallel load — 6/6 isolated); morfo:vocabulary exit 0; browser smoke of /uix/components/words green (sample renders, text + JSON import round-trip, invalid JSON rejected without mutating the doc). Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>	4 months ago
dev	5450d5f717	fix(words): gutter redesign — rail out, handle at block lower-left, "+" at block bottom-center ... Per spec: "la banda de la derecha la eliminamos, el drag aparece con el hover del block a la izquierda inferior y el insertar (+) en el centro cuando el hover este en el final del block, la line debe de ser mas solida sin blur" Layout: - Removed the dedicated rail grid column entirely. The editor host grid drops from 4 columns (rail | gap | content | drawer) to 2 columns (content | drawer), or (drawer | content) when side=left. - Removed `--_words-rail-bg/border/px/gap` CSS vars + the `::before` pseudo-element that drew the silver band. The rail recipe tokens (words.rail-bg / rail-border) become dead and can be removed in a later sweep; left for now to avoid generated-css test regen in the same commit. Block handle (drag grip): - Position math simplified: lives at the BLOCK'S lower-left corner with a 4px inset. `top = rect.bottom - 28 - 4`, `left = rect.left + 4`. No more rail-relative math, no detectDrawerSide / cssToPx / rememberRailMetrics helpers — all removed. - CSS base opacity bumped from 0.75 → 1 (the grip is shown via JS only on block hover; the visibility is binary, not faded). Block inserter ("+"): - findSeamForCursor rewritten: only triggers when the cursor is INSIDE a top-level block AND in its bottom portion (last 25% of block height, capped at 24px). When triggered, the seam anchors at the block's bottom edge spanning the block's full width — CSS centers the button via left:50% + translateX(-50%). - Removed listBlockBoundaries + SEAM_RANGE_PX (dead code under the new model — we don't compute inter-block seams anymore, we react to the hovered block's bottom). - Removed drawerSide flip + data-side attr (the "+" lives INSIDE the block now, not in a side gutter — drawer side is irrelevant). Line visual: - Removed transparency: `background: var(--_words-accent-solid)` (was color-mix(... 35%, transparent)). Opacity 0.9 always when shown (no opacity-0 default + hover transition). - Removed box-shadow halo on drop-mode state (was the blur effect the spec calls out). Drop mode still grows the line block-size from 2px to 3px so the drop target is unambiguous. Verification: DOM probe at /uix/components/words confirms: - Hover middle of block: handle visible at lower-left (top=666 vs rect.bottom=698, left=66 vs rect.left=62), inserter hidden. - Hover bottom of block: handle still at lower-left, inserter visible spanning block width (left=block.left, width=block.width), CSS centers the button horizontally. - `npm run check`: 0 errors. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	74e54384ea	feat(words): R4A.1 — foundation V2 ops (types + normalize + setBlockVisual + 5 block ops) + 31 tests ... R4A.1 opens the operations layer of the V2 engine. Pure functions only; no provider coupling. Lives in `engine/operations-v2/`. The full V1 op surface (~50 functions, ~3000 lines) splits across multiple R4A.* sprints — this one establishes the foundation + the V2-exclusive visual operation that unlocks POLISH-1b in later sprints. operations-v2/types.ts (~55 lines): - WordsEditorStateV2 = { document, selection, activeMarks } over V2 shapes. Reuses V1's `WordsSelection` (path + offset) verbatim — the selection model doesn't change between V1 and V2. - WordsOperationResultV2 = { state, changed }. Same shape as V1's result so future provider migration can plug in directly. - noOp + changed helpers for guard-clause and result construction. operations-v2/normalize.ts (~250 lines): - normalizeDocumentV2(doc, opts?) — defensive shape repair after every operation. Returns same document REFERENCE when nothing changed (cheap structural compare via mutated flag). - Autogen `id` for blocks/items/rows/cells without one (delegates to defaultIdGenerator from migrate-v2 → crypto.randomUUID + fallback). - Seeds empty containers: empty document → 1 paragraph; empty list → 1 item; empty table → 1×1 cell; empty inline children → 1 empty text. - Coalesces adjacent text inlines with identical marks (V1 invariant preserved — keeps the model from fragmenting after edits). - Strips marks from code block children (code is plain text only). operations-v2/visual.ts (~115 lines): - setBlockVisual(state, blockIndex, patch) — V2-EXCLUSIVE. - Shallow-merges patch into block.visual. Defends: * blockIndex must be valid. * Every patch key must be in WORDS_VISUAL_KEYS_PER_TYPE[block.type] (P8 whitelist enforced at runtime; unknown keys silently dropped). * Patch values of `undefined` REMOVE the corresponding visual property from the merged result (clears via undefined). * Empty visual is omitted from the block (no `visual: {}` stored). * No-op result when patch matches existing visual (avoids spurious re-renders). - This is the operation POLISH-1b (image radius slider + shadow toggles + border controls) will call. operations-v2/block.ts (~140 lines): - updateBlockAt(state, blockIndex, patch) — shallow-merges block fields, refuses to change `type`. - deleteBlockAt(state, blockIndex) — removes block; normalizer seeds an empty paragraph if it was the last. - insertBlockAt(state, blockIndex, block) — inserts at index (clamps to bounds). - moveBlockAt(state, blockIndex, direction) — swap with neighbor. - duplicateBlockAt(state, blockIndex) — inserts a copy with a fresh autogen id (drops the original's id before normalize so they're distinguishable). - moveBlockToAt(state, fromIndex, toIndex) — arbitrary repositioning for drag-drop. operations-v2/index.ts: barrel export of the public surface. Tests (31): normalize (autogen ids preserved when present, empty containers seeded for doc/list/table/inline children, text coalescing with marks-aware preservation, code stripping, idempotence via same reference). setBlockVisual (set/merge/clear/empty-elision/per-type whitelist enforcement/shadow on image vs paragraph/no-op on match). Block ops (updateBlockAt with type-change refusal + invalid index no-op; deleteBlockAt with seed-on-empty; insertBlockAt with clamping; moveBlockAt with edge no-op; duplicateBlockAt with fresh id; moveBlockToAt with from-equals-to no-op). Two TS errors landed mid-write (over-clever readonly modifier on function param + structural cast across discriminated union) — both fixed with simpler types. Verification: 257/257 engine tests pass (85 V1 + 50 V2 validator + 35 V2 migrator/parity + 56 V2 render/ serializers + 31 V2 ops). `npm run check`: 0 errors. Next: R4A.2 — text + selection operations (insertText, deleteRange, insertParagraph, deleteBackward, deleteForward, toggleMark, etc.). ~2500 lines of V1 to mechanically translate. Or pause R4 entirely and tackle the pragmatic shortcut (backport visual to V1 image to ship POLISH-1b sooner). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	194ca208f9	feat(words): R3 — render-v2 + HTML + MD serializers + 56 tests ... R3 closes the pure-V2 output path. Take a WordsDocumentV2 and turn it into either an abstract render tree (for Eidos/React/Vue/vanilla to walk), an HTML string with inline styles (roundtrip-friendly), or a Markdown string (lossy by design — visual.* dropped). engine/render-v2.ts (~370 lines): - Pure renderWordsDocumentV2(doc) → readonly WordsRenderElementV2[]. - WordsRenderNodeV2 is a superset of V1's WordsRenderNode (adds 'hr', 'div', 'section' tags for divider / callout). V1 stays untouched. - Per-block dispatch for all 9 V2 types (paragraph/heading/quote/code/ list/table/image/divider/callout). Same node shape as V1 — Svelte walker doesn't need changes. - visualToStyle(visual) helper exports the central translation from BlockVisual props to CSS declarations. Numbers get 'px', hex strings pass through. borderColor without borderWidth defaults to 1px solid. - Mark wrapping: parametric color/background marks combine into one <span style="...">; boolean marks stack via <code>/<span> with data-words-mark attr + inline style fallback. - Callout renders as <div role="note" data-words-callout-intent="X"> with optional title block + nested children. engine/serialize-html-v2.ts (~120 lines): - Pure serializeHtmlV2(doc, { pretty? }) → string. - Walks renderWordsDocumentV2 output and emits escaped HTML. - Properly escapes &, <, > in text; &, <, >, " in attribute values. - Void elements (img/hr/br/input/meta/link) emit self-closing form. - Pretty mode (opt-in) indents nested elements with 2 spaces. - Roundtrip-friendly: inline style="..." preserves visual.* lossless. engine/serialize-markdown-v2.ts (~190 lines): - Pure serializeMarkdownV2(doc) → string. - LOSSY by design (D-Q2): visual.* dropped entirely; color/background marks dropped (no MD vocabulary); underline mark dropped. - Preserves structural semantics: headings (# ## ###), quotes (>), code blocks (``` with language fence), lists (- / 1. / - [x] - [ ] GFM task lists), tables (GFM with :--- / :---: / ---: alignment), images (), dividers (---), callouts (GFM admonitions via INTENT_TO_GFM_ADMONITION mapping). - INTENT_TO_GFM_ADMONITION exported: P7 boundary mapping — neutral→NOTE, affirm→TIP, fulfill→IMPORTANT, risk→WARNING, threat→CAUTION, loss→CAUTION. - Mark order in output: bold(**) outer → italic(*) → strike(~~) → code(`) inner. - Cells escape | as \| to avoid GFM table parser confusion. Tests (56 new): - render-v2.test.ts (23 tests): basic blocks, visual sidecar with exact style strings, table headerRow/headerCol semantics, marks rendering, path encoding, visualToStyle direct. - serialize-html-v2.test.ts (12 tests): paragraph, escape semantics, void elements, inline styles, callout role attr, pretty mode, table th/td. - serialize-markdown-v2.test.ts (21 tests): every block type, marks, visual.* dropped (lossy), callout intent → admonition tag mapping (all 6 SemaIntent values), link with title, table cell | escaping. Important deferral: POLISH-1b (image radius/shadow/border slider) needs the provider to write block.visual.*. Provider is V1 internally. Deferred to R5 (after R4 migrates provider to V2). Pure render path is ready when provider catches up. Verification: 226/226 engine tests pass (85 V1 + 50 V2 validator + 35 V2 migrator/parity + 56 V2 render/serializers). `npm run check`: 0 errors. V1 model + render path UNTOUCHED — V2 lives in parallel. Next: R4 — provider V2 migration. Operations rewritten over WordsBlockV2. Runtime imageStatus sidecar wired at mount. Public API preserved via adapter during transition. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	6c70699ff7	feat(words): R2 — V1→V2 migrator + sema parity check + 35 tests ... R2 closes the engine refactor non-breaking phase. The V2 model from R1 is now reachable from any V1 document via a pure migrator, and the engine's local SemaIntent copy stays in lockstep with the canonical $uix/intent.INTENTS via a runtime assertion. engine/migrate-v2.ts (~370 lines): - Pure function migrateV1ToV2(input, opts?) → { document, runtimeState, warnings }. - Per-block migration paths for paragraph / heading / quote / code / list / table / image / divider / callout. Each rejects invalid values defensively (e.g. invalid heading level coerced to 1 with warning). - P4 enforcement on image: drops `image.status` from the content, moves it to `runtimeState.imageStatus: Map<blockId, status>` keyed by an autogen id. Caller (provider) restores into in-memory sidecar on load. - P1 sweep on table: drops `striped` / `compact` (presentation tokens) with warnings. Drops `cell.tone` (design-system token) with warning. Renames list.children → list.items silently. - Marks reshape: V1 template-literal `'color:#hex'` / `'bgcolor:#hex'` → V2 structured `{type:'color'|'background', value:'#hex'}`. Already-structured marks pass through unchanged. - Code blocks narrow to text-only: flattens any link inside code into its text content with warning. - Idempotent on V2 input: a V2 document round-trips with zero warnings and exact-equal output (validates with validateWordsDocument). - Default id generator uses crypto.randomUUID() when available, with Math.random + monotonic counter fallback. Tests inject a deterministic sequential generator. engine/sema-parity.ts (~60 lines): - assertSemaIntentParity(canonicalIntents) throws SemaIntentParityError if the engine's local WORDS_EVAL_INTENTS drifts from the canonical list. Error names both the missing and extra values + points to the file to update. - Caller passes canonical list (avoids cross-package import in this module). Provider mount calls it with $uix/intent.INTENTS. engine/migrate-v2.test.ts (35 tests): - Entry-point shape handling (null / non-object / non-array children). - Per-block migration (paragraph, heading with level clamp, quote with cite, code with text-only narrowing + link flattening, list with kind coercion and indent-0 omission, table with token drops, image with src validation). - P4 enforcement: image.status → runtimeState.imageStatus keyed by autogen id; respects caller-provided id. - Marks reshape: template-literal → structured; structured pass-through; invalid hex rejected; unknown boolean dropped. - Idempotence on V2 input (paragraph, callout, divider). - Callout intent: all 6 canonical SemaIntent values accepted; invalid (e.g. 'warning') coerced to 'neutral' with warning. - Post-condition: migrator output ALWAYS passes validateWordsDocument on any V1 input. - Sema parity: 5 tests including the real golden cross-check against $uix/intent.INTENTS. Verification: 170/170 engine tests pass (85 V1 + 50 V2 validator + 35 V2 migrator/parity). `npm run check`: 0 errors. Next: R3 — visual sidecar implementation. Render emits inline styles from block.visual.*; POLISH-1b (image radius/shadow/border controls) becomes implementable under this model. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	9b13e3cf23	feat(words): R1 — V2 architecture proposal + types-v2 + validator + 50 tests ... Architectural refactor proposal landed for review. R1 closes the audit + type design + validator phase. V1 stays untouched in parallel. ARCHITECTURE_PROPOSAL.md (~500 lines): - 8 principles signed (P1 content↔design strict, P2 raw values only, P3 renderer translates, P4 runtime state out, P5 framework-neutral engine, P6 single source of truth, P7 sema intents canonical for evaluative vocabulary, P8 per-property entry rules). - D-SEM-1/2/3 + Q1-Q6 all signed with recommendations accepted by user. Notable: CalloutBlock.intent uses SemaIntent canonical (neutral|affirm|fulfill|risk|threat|loss), NOT web admonition vocabulary (warning|info|danger|etc.) — those live as UI labels + Markdown serializer mappings ONLY. engine/types-v2.ts (~320 lines): - WordsDocumentV2 with version '2.0.0'. - BlockBase with optional id (autogen at load) + optional visual sidecar. - BlockVisual with RAW values only (numbers in px, hex color strings). Per-type Pick<> subsets: ParagraphVisual / HeadingVisual / QuoteVisual / CodeVisual / ListVisual / TableVisual / ImageVisual / DividerVisual / CalloutVisual. - 9 block types: paragraph, heading, quote, code, list, table, image PLUS new divider (semantic <hr>) and callout (with WordsEvalIntent + nestable block children). - Marks refactored from V1 template-literal `color:${string}` to structured `{type:'color'|'background', value:'#hex'}`. Type-safe. - Image upload status REMOVED from the content model (P4): the WordsImageBlockV2 no longer has `status`; it goes to runtime sidecar `provider.runtime.imageStatus: Map<blockId, status>`. - Table V2 drops `striped` + `compact` (presentation tokens). Adds `headerRow` + `headerCol` (semantic accessibility). Per-cell and per-row `visual` allowed (background for zebra; cell padding). - Cell `tone: 'muted'|'accent'` REMOVED (was a design system token). Replace with cell.visual.background hex if user wants a tinted cell. - WordsEvalIntent declared locally in engine to keep engine framework-neutral. Parity with $uix/sema asserted at boot (TODO in R2). engine/validate-v2.ts (~520 lines): - Pure function validateWordsDocument(input) → ValidationResult. - 11 error codes: invalid-version, invalid-block-type, invalid-block-shape, invalid-children, invalid-visual-key, invalid-visual-value, invalid-mark, invalid-href, invalid-eval-intent, invalid-enum-value, invalid-number, duplicate-id, invalid-text-content, invalid-link-children. - Enforces P2: hex regex on color fields, finite numbers on dimension fields, rejects tokens / classes / var() refs / non- canonical mark shapes. - Enforces P8: per-type Pick<> at runtime via WORDS_VISUAL_KEYS_PER_TYPE whitelist. - Enforces P7: callout.intent must be one of the 6 canonical sema intents; rejects web vocabulary (`warning`/`info`/etc.). - Each error carries a JSON-pointer path (e.g. /children/1/visual/background). engine/validate-v2.test.ts (50 tests): - Top-level shape, block type registry, marks (boolean + structured + V1 reject + token reject), visual properties (allowed + rejected tokens/classes/var()/non-numeric + per-type Pick<>), id uniqueness, heading levels, list kind + indent, code text-only children, image src + width, callout SemaIntent + reject of web vocabulary + nested blocks, divider, quote.cite URL parseability, link nesting + URL parseability, table headerRow/headerCol + per-cell visual whitelist + per-row visual. Verification: 135/135 engine tests pass (85 V1 + 50 V2). V1 untouched. npm run check: 0 errors, 26 warnings (pre-existing). Next: R2 — migrator V1→V2, runtime image-status sidecar, refactor provider to consume V2 internally. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	d126e16e65	fix(words): silence canvas truly, drag for list/table, no overlay hovers + day-end docs ... EV-G final pass + EOD documentation. Words editor: - Stop emitting `contact-focus` from soma. Removing the per-component cascade rule earlier didn't fully silence the editor — the `contact` family carries a BASE sound signature in SEMA_MAP (pitch 800, gain 0.25) that fires even without a cascade rule. The trigger call is commented out (event no longer emitted; telemetry empty). Per spec: "sigue apareciendo el puto sonido, el ring". - Drag handle + inserter now cover lists AND tables. Selectors expanded from `[data-words-node="block"][data-words-path]` to also match `[data-words-node="list"]` and `[data-words-node="table"]`. Top-level lists / tables get the grip + the "+" seam under them. - Removed :hover from `[data-words-block-handle]` and `[data-words-block-inserter-button]`. Gutter overlays are AMBIENT — visual state via data-attrs only (`data-open`, `data-grabbed`, `data-dragging`). Per spec: "quítale el hover al botón de añadir bloque, si el handle no tiene este tampoco". - Rail tokens hoisted to recipe: `words.rail-bg = #d4d4d4`, `words.rail-border = #9a9a9a`. Fixed-tone (not theme-aware) — the rail emulates the physical margin of a notebook; same look in light/dark. Recipe-css-contract test was rejecting raw hex in words.css. Docs: - CLAUDE.md hand-off 2026-05-28 #2 documenting the 6 EV-* Words sprints + 5 architectural decisions with lasting effect (canvas silent, fixed-tone tokens for physical metaphors, drag predicate covers list/table, ambient overlays, insertParagraph empty-block guard). - continue.md (CONTINUE.md) rewritten for tomorrow's pickup: TL;DR of the two halves (toggle-group structural + Words UX), remaining bugs (drag handle outside rail on code blocks, heading inline level change), and the heredados (POLISH-1b, F2.3-F2.6, DRAGDROP-SOMA). Verification: 467/467 morfo + sema + eidos + soma/words tests pass. 85/85 engine pass. `npm run check`: 0 errors. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	125c62cc20	fix(words): silence canvas, silver rail, direct undo/redo, popover widths, Enter empty ... Six user-reported issues in one pass. EV-F1 — canvas silence Remove the contact-focus entry from the sema cascade. The event still fires (telemetry / a11y can read the data-event* stamps) but no sound / haptic plays. Per spec: "no puede emitir eventos es demencial, una vez que está trabajando con el no tiene sentido". The editor surface is now perceptually silent on focus; only explicit commits (save, format, set-link, etc.) make noise. EV-F2 — rail visual `--_words-rail-bg` becomes flat silver `#d4d4d4` and the dot pattern is gone. New `--_words-rail-border` (`#9a9a9a`) lays a 1px hairline on the rail's inline-end so it reads as a separate gutter ("paper edge" feel). Fixed-tone (not theme-aware) — the rail is meant to look like the physical margin of a notebook regardless of the editor theme. Left-side drawer flips the border to inline- start. EV-F3 — direct undo/redo in toolbar Demo presets + custom config now use `['undo', 'redo', ...]` as direct command strings instead of `['history-menu', ...]` (which opened a popover to pick from two items — wasted real estate + two clicks). `link-menu` and `tools-menu` stay as popovers because they hold 2-3 items each. `insert-menu` is OUT — block construction lives in the drawer's empty-state Insert panel and the slash menu. `demoToolbarFamilyItems.insert` collapsed to `[]` accordingly. EV-F4 — popover widths Base `[data-words-toolbar-family-panel]` width 12.5rem → 16rem so content fits without horizontal scroll. Per-family overrides: tools → 17rem (descriptive labels), link → 20rem (URL input + button grid). The default panel CSS no longer carries a max-block- size + scroll — a vertical scroll inside these popovers is a misdesign signal; size the panel, don't scroll it. EV-F5 — link popover (covered by EV-F4) The vertical scroll the user saw was a side-effect of the 12.5rem width forcing the link-editor input + button to wrap, then the panel running out of vertical room. The new 20rem width restores the natural one-row layout and no scroll appears. EV-F6 — Enter on empty block no longer multiplies `insertParagraph` engine op now guards the default case (paragraph/heading): if the block is collapsed AND empty (children length 0, both halves of split empty), Enter on a paragraph is a no-op and Enter on a heading demotes it to a paragraph (canonical Notion / Google Docs UX). For non-empty paragraphs/headings the existing split-and-create-paragraph behavior stands. Verification: 467/467 tests pass across morfo + sema + eidos + soma/components/words. 85/85 engine tests pass (insertParagraph guard non-regressive on existing fixtures). `npm run check` still 0 errors. DOM probe at /uix/components/words confirms toolbar = 5 items (undo, redo, Links, Tools, Find/replace), rail bg `rgb(212,212,212)` + 1px `rgb(154,154,154)` right border, link popover 320px wide with no scroll. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	ce95473261	feat(words): internal scroll for the content area + capture-phase scroll sync ... The editor body was unbounded vertically — long documents pushed the whole page instead of confining the scroll to the editor surface. Per spec: "la parte de la edicion deberia tener un scroll". Recipe: - Three new tokens `content-max-block-size-sm/md/lg` (50vh / 60vh / 70vh) mirror the existing `content-min-block-size-*`. Generated CSS regenerated. CSS: - `[data-words-content]` now carries `max-block-size: var(--_words-content-max-block-size)` and `overflow-y: auto`. The min-block-size baseline stays as the starting height for empty editors. When content grows past the cap, the body scrolls internally — the host card no longer balloons. - `scroll-padding-block: 1rem` hint so caret-into-view nudges leave breathing room. Block-handle / block-inserter / image-float-bar: - Switch scroll listeners from `window.addEventListener('scroll')` to `document.addEventListener('scroll', { capture: true })`. Scroll events don't bubble; the previous wiring only caught window scrolls. Now the gutter overlays (handle position, "+" snap target, image float bar anchor) re-measure when the user scrolls INSIDE the content area too. Verification: dev probe at /uix/components/words confirms `max-block-size: 600px`, `overflow-y: auto`, content scrollHeight (615px) > clientHeight (600px), `isInternallyScrollable: true` on a 1000px viewport (60vh of 1000 = 600). 467/467 tests pass across eidos + morfo + sema + soma/components/words. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	c4be6dbe1b	fix(words): silence body event spam + slim top toolbar + reposition handle/inserter ... Four user-requested fixes to the Words editor surface — the body should not emit perceptual cues during editing, the top toolbar must not duplicate the drawer, and the gutter widgets need clearer anchor points. EV-A (events spam): - Move `contact-focus` event target from `content` to `provider` in the morfo + sema cascade selector. Focus is a Words-component-level signal; firing it from the content body conflates "user entered the editor" with "user is editing the body". The wasFocused guard in soma already throttles to one emit per real focus transition. - Expand `isInsideWordsTool` selector list to cover the four companion overlays added after the original list was written: `data-words-drawer`, `data-words-block-handle`, `data-words-block-handle-menu`, `data-words-block-inserter`, `data-words-image-float-bar`. Before this fix every click on a companion overlay fired commit-save-content + contact-focus on the blur/refocus cycle (two perceptual cues per tool interaction). EV-B (toolbar trim): - Demo's full / formatting / minimal presets + custom config now host only GLOBAL actions: history (undo/redo), insert (creates new blocks), link (selection-bound flow), tools (find/replace + clear), find-replace in its own group. text-menu / block-menu / align-menu / list-menu / table-menu moved out because the drawer already owns contextual formatting per the HIER-2 split. EV-C (block-handle drag UX): - Drop the `setDragImage(hoverBlockEl)` call. The browser now uses its default snapshot (the grip button itself) as the drag ghost — the ghost travels with the cursor while the bar in the gutter stays fixed as a visual anchor. New `data-dragging` attr + `[data-words-block-handle][data-dragging]` CSS rule fades the static anchor to 0.35 opacity so it reads as "drag origin" while the ghost is the moving part. EV-D (inserter at block bottom): - Seam positions are now pinned to the BOTTOM EDGE of the preceding block (`a.bottom`) instead of the midpoint of the gap between two blocks. The "+" reads as "insert AFTER this block" anchored to that block's lower edge, per spec — el botón de añadir bloque debe aparecer en el límite inferior del área en relación al bloque. - Tighten the "cursor inside block band" check to a half-open interval `[top, bottom)` so the exact bottom-edge pixel belongs to the seam below (otherwise the seam at `y === bottom` is shadowed by the block and the inserter never snaps). Verification: dev server, /uix/components/words, DOM probe confirms handle centered in the rail column (left=48 inside the 32-wide rail starting at ~46), inserter snaps to block 1's bottom (`top=569.94px` when block 1 bottom = 570px). 366/366 tests pass in morfo + sema + soma/components/words. `npm run check` still 0 errors. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	c250f7d790	fix(words): close 12 type-drift errors from F2/F3/COLOR sprints ... After the toggle-group structural migration `npm run check` still had 12 pre-existing errors. All from the active Words sprint (post-2026-05 F2/F3/COLOR work), not introduced by this session. Closes them to restore green check: - soma words.svelte: thread `onUploadImage` through the Provider opts (was required-undefined in WordsOpts but missing from the create() call). - eidos words-toolbar-family-menu.svelte + words.svelte: add 'insert-image' label to COMMAND_LABELS (the Record<WordsToolbarButton CommandName, string> required full coverage). - eidos words.svelte: hoist the `leafItem` snippet out of <SomaWords. Provider> to template root level. Snippets defined INSIDE a component element are passed as named props in Svelte 5; this one was meant as a local helper. Also kills the "Expression too complex" error at the Provider tag — the leafItem prop was bloating the union. - eidos words-drawer.svelte: extract 5 inline button-config arrays to typed `as const satisfies readonly { id: WordsCommandName; ... }[]` constants (FORMAT_MARKS, TEXT_ALIGN_ACTIONS, CELL_TONE_ACTIONS, CELL_VERTICAL_ACTIONS, CELL_TEXT_ALIGN_ACTIONS). The inline literals were inferring `id: string`, breaking `runCommand` and `activeMarks.includes`. Result: npm run check 16 → 0 errors. Tests 658/658 pass. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	d2f184da44	refactor(toggle-group): item is structurally a Toggle (zero derivation duplication) ... Closes the deuda left explicit in hand-off 2026-05-27 #5: the inlined derivation expressions in toggle-group.css duplicated Toggle's recipe because [data-toggle-group-item] wasn't a descendant of [data-toggle]. Resolution is structural, not a drift test: - morfo: each item declares { attr: 'data-toggle', value: v.literal(''), severity: 'required' }. Captures structural identity — an item IS a toggle in every observable sense (same press, same variant/color/size, same on/off machine). - eidos: new context.ts propagates the root's variant + size via reactive getters; the item wrapper writes them as data-attrs on its button. Combined with data-toggle, the item is DOM-equivalent to a standalone <Toggle>. - eidos css: ~150 lines deleted (base derived tokens, variant cascades, size cascade, focus/disabled/icon-only duplicates). Toggle's recipe now paints the item end-to-end. CSS keeps only the grouping concerns: flex, orientation, attached, block, focus z-index, group disabled. - composition (TSC v2.2) stays — palette overrides land on the item under [data-toggle-group][data-color='X'] [data-toggle-group-item], which Toggle's --toggle-palette-* chain reads at item scope. Verification: - DOM probe (12 combinations: 4 colors × 3 variants × 2 states) on /uix/components/toggle-group. Computed values match bit-a-bit with baseline pre-refactor (e.g. affirm/solid on = rgb(18,165,148), risk/outline on = srgb(0.2,0.118,0.043), threat/ghost on = rgb(25,17,17)). - vitest src/uix/{morfo,eidos}: 163/163 pass. - vitest src/uix/soma/components/toggle-group: 4/4 pass. - npm run check: 12 errors (all pre-existing in active Words sprint; zero added by this migration). Drive-by: the same npm-check pass surfaced 4 stale 'orientation' refs in words-toolbar.svelte + words-toolbar-group.svelte from this session's earlier toolbar refactor (orientation was removed from types but not from these soma components). Removed; brings check from 16 → 12 errors. CLAUDE.md hand-off 2026-05-28 documents the new doctrine: when a wrapper visually reuses another, declare structural identity in the wrapper's morfo. Don't duplicate the cascade, don't extend TSC for one case. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	949cc88d75	fix(codex-audit): narrow content non-null in block-handle + inserter ... `npm run check` was reporting 'content is possibly null' errors in both block-handle and block-inserter $effect closures — 6 errors between them. TS doesn't carry the `if (!content) return` narrowing into inner closures, so we re-bind to a const typed `HTMLElement` right after the guard. `npm run check`: 22 → 16 errors. Remaining 10 (drawer command-name / mark string-not-assignable, words.svelte missing 'insert-image' key in the Record, plus 6 pre-existing in src/lib/_demo and others) need a deeper look at the command-name / mark string unions — follow-up. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	d767d88c03	fix(codex-audit): close P0 + tokenize drift introduced by Words sprint ... Three categories of audit drift introduced during the recent Words sprints (drawer redesign, image float bar, color marks, insert panel, gutter rail), all caught by the existing contract tests now that the work has landed: 1. **Undeclared morfo attr** (P0 contracts test): `data-words-block-selected` is stamped by the engine's render on the atomic-selected block (today only image) but wasn't in the morfo's data-attr list. Adds it as optional on the same parts array that already declares `data-words-image-*`. 2. **Missing i18n key** (translations:check): The drawer part references `components.words.drawer` for its aria-label, but the langs catalog was never updated when DRAWER-CANON landed. Add `drawer` to `src/uix/langs/components/words.ts` (es + en). 3. **Recipe CSS contract drift** (eidos recipe-css-contract test): - 3 undeclared CSS vars used in the drawer-Editable styles: `--words-toolbar-color` → `--words-command-color` `--words-command-hover-bg` → `--words-command-bg-hover` `--words-command-active-color` → `--words-command-color-hover` (Right names per `lib/recipes/base.ts > words`.) - 15 raw font-size literals (0.65rem / 0.7rem / 0.75rem / 0.8rem / 0.9rem) replaced with the existing words font-size tokens: 0.65/0.7rem → `var(--words-status-font-size)` 0.75/0.8rem → `var(--words-font-size-sm)` 0.9rem → `var(--words-font-size-md)` Result: - `contracts.test.ts`: 31→32 pass (the data-attr violation is gone). - `translations:check`: 219 refs / 76 catalogs / 0 errors / 0 warnings. - `recipe-css-contract.test.ts`: 17→19 pass. - Full `vitest run src/uix`: **867/867** pass (was 865/2-fail before). - `npm run check` clean (only 6 pre-existing errors unrelated). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	08ae867caf	fix(words): grip lands in the rail (rem→px) + drag preview is the block ... Two bugs blocking the drag-and-drop UX: 1. **Grip was 22px off** — `parseFloat('2rem')` returns 2, not 32. I was reading `--_words-rail-px` / `--_words-rail-gap` as raw strings and treating them as pixels. With the actual rem values the grip landed in the gap/content padding instead of the rail. Fix: `cssToPx(raw, rootFontSize)` resolves `rem`/`em` → px via the documentElement's computed font-size. With the fix the grip's center matches the rail's center exactly. 2. **Drag preview was the tiny grip icon, not the block** — native HTML5 drag defaults to a snapshot of the `draggable` element. Since `draggable=true` lives on the 28×28 grip button, the user saw a tiny grip ghosted under the cursor instead of the paragraph/heading/image being moved. Fix: `e.dataTransfer.setDragImage(hoverBlockEl, 12, 12)` in `ondragstart`. The browser now uses the BLOCK's element as the drag image — user sees the whole row/heading/image moving with the cursor, just like Notion / Linear / Figma. Note re: soma's `<DragDrop.Provider>`: the user is right that soma has a first-class drag-drop system (pointer + keyboard + ARIA). The reason this implementation still uses native HTML5 dragstart/ dragend is architectural: soma's `<DragDrop.Draggable>` is designed to WRAP each draggable item, but here the blocks are rendered by the engine's HTML render function as plain DOM (not Svelte components), so each block isn't wrappable. The block-handle is a single floating overlay that snaps to whichever block is hovered. Migrating to soma DragDrop would require restructuring the editor render to per-block Svelte components — a larger architectural change. Filed as a follow-up. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	ccd9d3e0ab	fix(words): "+" inserter only shows in the rail gap between blocks ... Previously the "+" snapped to any seam within 18 px of the cursor's Y — which meant it appeared even when the cursor was hovering on a block's body (near its top/bottom edge) or in the rail aligned with the block. That clashed with the grip's territory. Per user spec: when the cursor is on a block (text or rail-aligned), the grip is the actor — the "+" should be hidden. The "+" only surfaces when the cursor is in the rail column AND vertically positioned in a gap BETWEEN two blocks. Two new gates added to `findSeamForCursor`: 1. **Out-of-content X**: if `e.clientX` falls inside the content box's horizontal extent (the editor's text area), return null immediately. The cursor is reading text or near the toolbar/ status row, not the rail. 2. **Out-of-block Y**: if `e.clientY` falls inside ANY top-level block's vertical band, return null. The block-handle owns that slot. Only when both gates pass do we snap to the nearest seam. The existing `dragover` path is untouched — drag indicator still works inside the content area because the drag operation has different ergonomics (you're actively moving a block, the line is the drop target, not a discoverability cue). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	6bcbd8a983	refactor(words): rail is now a real column outside the editor box ... User feedback was right — the previous approach put the rail INSIDE the editor as extra content padding, so it was glued to the editor card with no visual separation. Reframed as a proper SIBLING column in the grid layout, with its own grey/dotted background and a gap to the editor mirroring the drawer's gap on the other side. Layout (right-drawer default): [rail 2rem] [gap 0.5rem] [editor card] [drawer] Left-drawer flips both rail + gap to the right side. CSS changes: - New tokens: `--_words-rail-px` (2rem ~32px column width), `--_words-rail-gap` (0.5rem ~8px gap to editor), `--_words-rail-bg` (subtle grey using toolbar-border tint), `--_words-rail-dot` (denser status-color dots at 65% opacity). - Grid `template-columns` and `template-areas` of `[data-words]:has([data-words-drawer])` extended to include `rail` and a `.` gap column. Left-drawer mirror reverses them. - Rail visual: `[data-words]:has(...)::before` with `grid-area: rail`, dotted radial-gradient on grey bg, rounded corners matching the editor's radius. - Reverted content + placeholder to symmetric `padding-inline` (the rail no longer lives in content's padding). Deleted the obsolete `[data-words-content]::before` rule. Overlay positioning (JS): - `words-block-handle.svelte` reads `--_words-rail-px`, `--_words-rail-gap`, `--_words-content-px` from the editor host and computes the grip's `left` as the center of the rail column in viewport coords: right-drawer: `block.rect.left - content-px - rail-gap - rail-px/2` left-drawer: `block.rect.right + content-px + rail-gap + rail-px/2` Then offset by half the grip's width (14 px) to center on the rail. Inserter "+" button (CSS-only): - `left` switched from a hard `-28px` to a calc that mirrors the grip math: `-(content-px + rail-gap + rail-px/2) - 12px` Right-edge mirror for the left-drawer case. Grip visibility bumped: - Color from `status-color` to `command-color` so it reads as active-text against the dotted rail. - Default opacity 0.45 → 0.75 (visible without hover). - Hover state gains a toolbar-bg fill + accent-text + accent border ring so the hovered grip stands out from the rail. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	33b3a580b7	feat(words): visible dotted band in the grip rail + trim to 1rem ... Two complaints from the user on the rail: 1. Still too wide at 1.5rem (~24px). 2. The band doesn't visually distinguish itself from the canvas so the grip + "+" look detached. Fixes: - **Rail trimmed to 1rem (~16px)** — combined with content-px gives a 32px total padding on the rail side. The 28×28 grip lands at content.left+0 → +28 with a 4px gap before the text. This is the minimum band width that still keeps the grip fully inside the editor's visible content area. - **Dotted pattern on the band** — `[data-words-content]::before` fills the rail's padding zone (32px wide, full height) with a radial-gradient dot pattern at 8×8 px tiles. Color is `status-color @ 40% opacity` so the dots read as a "this is a gutter" cue without competing with content. `pointer-events: none` and `z-index: 0` so they don't interfere with selection or caret. - **Flips with drawer side** — when `<Words.Drawer side='left'>` the pseudo's inset moves to inline-end so the dotted band tracks the rail to the right side. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	0edb7be26c	fix(words): trim the gutter rail to 1.5rem (~24px) ... Previous 2.5rem (~40px) made the content's left padding total ~56px which felt excessive — the band was very wide and the grip looked lost in the middle of it. Trim to 1.5rem so the total padding is ~40px on the rail side: tight enough that the 28×28 grip occupies most of the band (content.left+8 → +36) with just enough breathing room (4 px) before the text starts. Same Notion-style density. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	5a3029e589	feat(words): real gutter rail for the block-handle + inserter ... The grip `::` and the "+" inserter were positioned at `block.rect.left - 28..32` — fine in theory but the content area had symmetric `padding-inline` so that math landed the overlays in empty body space outside the editor box. User-visible result: the left band had no width, the grip dragged on/off the edge unevenly, and the click target felt invisible. Fix: dedicate a visible rail of `2.5rem` (~40px) on the side OPPOSITE the drawer, by switching the content + placeholder from symmetric `padding-inline` to an asymmetric pair (`padding-inline-start` + `padding-inline-end`). New token `--_words-rail-px` is added to one of those sides per drawer position. - Drawer on RIGHT (default): rail is on the LEFT — content's `padding-inline-start` becomes `content-px + rail-px`. The grip at `block.rect.left - 32` now lands inside that visible band (≈ 24-52 px from the editor's left edge), with the inserter "+" button at -28 alongside it. Block text starts 56 px from the editor's left edge. - Drawer on LEFT: mirrored — the asymmetric padding flips via `[data-words]:has([data-words-drawer][data-side='left']) > [data-words-content]` so the rail moves to the inline-end side where the overlays now render. No JS changes needed — the existing geometry-based positioning (`block.rect.left - 32`, `seam.left - 28` for the "+", with the drawer-side detection added in an earlier commit) automatically lands in the new band because `block.rect.left` shifted right by the rail width. Verified in browser: hovering anywhere over a paragraph reveals the grip at the expected band position, with no flicker, and the hit area is comfortably wide for a click. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	377168550c	feat(words): drawer follows emptiness — Insert when empty, Inspect when filled ... User-clarified rule for the right drawer: when the current block is empty, the drawer shows ONLY the constructors (Insert panel) — no point in showing format/block-scope options on a blank line. When the block has any content, the inspector for that content type appears (paragraph + text-align, heading, image, code, etc.). Implementation: - New helper `isEmptyEditableTextBlock(block)` next to the slash command catalog. It returns true for paragraph / heading / quote / code-without-language whose single text child is the empty string. Atomic blocks (image / table / list) never qualify — they always carry structure. - `WordsDrawerProvider.modes` derivation gates the scope-stack computation behind `!isEmptyTextBlock(currentNode)`. Empty text blocks fall through directly to `['default']` (just Insert); blocks with content build the scope stack as before (format / block / cell / row / table / list-item / list) and Insert collapses at the bottom of the stack. - Atomic blocks (image / code with language) still push their scope unconditionally because their "content" lives in attributes, not children. Before: clicking "+" → empty paragraph inserted → caret on it → drawer showed `['block', 'default']` with Block expanded (text-align controls for an empty line — useless), Insert collapsed below. After: clicking "+" → empty paragraph inserted → caret on it → drawer shows `['default']` only (Insert expanded). As soon as the user types a character, Block panel appears above with Insert collapsed underneath. 152/152 soma words tests still pass; `npm run check` clean. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	2bf6aef035	fix(words): eliminate block-handle flicker in the gutter + bump hit area ... Two issues with the block-handle grip the user just hit: 1. Same flicker pattern the inserter had before its fix: the handle's listener was on `[data-words-content]`, and the grip sits OUTSIDE that box in the gutter. The moment the cursor moved toward the grip, `mouseleave` fired on content → grip vanished → cursor back inside content → grip reappeared → loop. 2. The 24×24 hit area sitting in a 28-px-wide gutter was hard to land, especially when the gutter also got the flicker treatment. Fixes: - **Document-level mousemove** (was: content-level). Bounds are checked manually against the content rect with an EXPANDED buffer on the side where the grip lives (56 px) and a tight 8 px on the opposite. Drawer side is read from the DOM so the bounds flip automatically with `<Words.Drawer side="left">`. - **Early-return when the cursor is over the handle or its menu** (`target.closest('[data-words-block-handle]' | '...-menu')`) so hovering the grip itself never recomputes/hides the state. - **Geometric fallback** when the cursor is in the gutter and `findBlockElement(e.target)` would return null: walk the top-level blocks and pick the one whose vertical band contains `e.clientY`. Keeps the grip pinned to the right block while the user is moving toward it. - **Mouseleave listener removed entirely** — the bounds check above handles the "cursor truly left the editor area" case without the flicker-prone per-element leave events. - **Hit area bumped to 28×28** (`1.75rem`) so it's comfortably above WCAG 2.5.5's 24×24 minimum and easier to grab in the narrow gutter. Visible icon stays at `size="sm"`. Centering math updated (-14 instead of -12); horizontal offset bumped from `-28` to `-32` to keep a small gap between grip and text descenders. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	4cd54e2111	feat(words): drawer `side` prop (left | right) + overlays auto-mirror ... The companion drawer was hard-coded to the right column. Now `<Words.Drawer side="left">` flips the editor grid so the drawer takes the start column; the floating overlays (block handle grip, block inserter "+") automatically render on the OPPOSITE margin so they don't crash with the drawer's gutter. Default stays `'right'` so existing consumers see no change. Wiring: - Eidos `<Words.Drawer>` gains `side?: 'left' | 'right'` (default 'right'). The value is passthrough'd as `data-side={side}` on the rendered drawer root so CSS + sibling overlays can read it from the DOM. - `words.css`: - `[data-words]:has([data-words-drawer][data-side='left'])` flips `grid-template-columns` + `grid-template-areas` so the drawer column sits at the inline-start. - `[data-words-drawer][data-side='left']` swaps the visual seam from `border-inline-start` to `border-inline-end`. - `[data-words-block-inserter][data-side='left'] [data-words-block-inserter-button]` overrides `left: -28px` to `right: -28px` so the "+" lands on the block's right margin. - Block-handle (`words-block-handle.svelte`): new `detectDrawerSide()` reads `[data-words-drawer]@data-side` once per measurement. The grip's `position.left` becomes `rect.right + 4` when drawer is on the left (was always `rect.left - 28`). - Block-inserter (`words-block-inserter.svelte`): same `$derived` read of the drawer's `data-side`, propagated as `data-side` on the inserter overlay element so the CSS rule above can flip the "+" button. - Demo: new `let drawerSide = $state<'left' | 'right'>('right')` + a chip control "drawer side: right | left". `<Words.Drawer side={drawerSide} />`. Verified `npm run check` clean (pre-existing errors only). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	3128b5f13c	refactor(words): drawer reinvented as Insert ↔ Inspect ... The right drawer's default panel is reframed: instead of an outline + stats overview that doesn't compose well with the editing surface, the default mode now shows an "Insert" catalogue of block constructors (Paragraph, Headings 1-3, Quote, Code, three list types, Image, Table). When there's contextual scope (image / table / code / format), those panels stack above with Insert collapsed at the bottom as a secondary affordance. Same Figma / Linear "right inspector" pattern but Notion-aware on the empty-context default. Architecture: - New `WordsDrawerMode` value: `'document'` (outline + stats + suggestions, opt-in via a header toggle). The old default content moved there verbatim. `'default'` was renamed (label-side) to "Insert". - `WordsDrawerProvider.modes` now ALWAYS appends `'default'` to the end of the stack. So: - No selection → `['default']` - Caret on paragraph → `['block', 'default']` - Inside table cell → `['cell', 'row', 'table', 'default']` - With doc panel toggled on → `['document', ...above]` Eidos auto-collapses the trailing `'default'` whenever any scope is also active, so it sits as a footer chip the user can expand to drop a new block without leaving the current inspector. - `documentPanelOpen = $state(false)` + `toggleDocumentPanel()` method on the provider. Exposed in the drawer snippet props as `documentPanelOpen` / `toggleDocumentPanel()`. - New provider method `insertBlockOfType(id: WordsSlashCommandId)`: reuses the slash-id space (paragraph / heading-* / quote / code-block / *-list / image / table) but with different semantics: - Empty paragraph context → `setBlock` (transform in-place, no stray blank). - Non-empty → insert a fresh block AFTER current via `insertBlock`. Caret lands at start of the new block. - Image / table → existing `insertImage` / `insertTable` commands (URL prompt still inline). Exposed in `ProviderSnippetProps` so the Insert panel buttons call it directly without going through the slash menu state. Eidos panels: - `insertPanel` snippet: a 2-column grid of `[data-words-drawer- insert-item]` buttons, each with icon + label. Click → `s.insertBlockOfType(item.id)`. - `documentPanel` snippet: outline + stats + suggestions (verbatim from the old default). - Header gains a circular `FileText` toggle on the right (`[data-words-drawer-document-toggle]`) — `data-active` when doc panel is in the stack. Same hover affordance as the rest of the drawer header. CSS additions: - `[data-words-drawer-insert]` grid layout. - `[data-words-drawer-insert-item]` button styling — neutral border, accent on hover. - `[data-words-drawer-document-toggle]` header chip with active state. Verified: 152/152 soma words tests pass. `npm run check` clean (only pre-existing errors). Browser smoke-test confirmed clicking "Heading 1" with caret on the image block inserts a new heading after the image and switches drawer to Block scope, Insert collapsed below. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	71c24dca9b	refactor(words): drop vertical toolbar orientation ... The companion drawer + new floating overlays (image bar, block handle, block inserter) assume a horizontal toolbar above the content. The vertical layout was a CSS-grid two-column repositioning that fights that geometry and was the cause of the new design breaking when the user switched to vertical. `horizontal` is the only orientation now. Cuts: Morfo (`morfo/components/words.ts`) - Toolbar + ToolbarGroup parts: `data-orientation` becomes a literal `'horizontal'` (no longer reads from `propRef('orientation')`). - Same for the toolbar's `aria-orientation`. Soma (`soma/components/words/types.ts`, `exports.ts`, provider) - Removed `WordsToolbarOrientation` type entirely. - Removed `orientation?` from WordsToolbarProps and WordsToolbarGroupProps. - Provider opts no longer read or pass through `orientation`. Eidos (`eidos/components/words/{types,words.svelte,words-toolbar.svelte,words.css}`) - `WordsProps.toolbarOrientation` removed. - `<Words.Toolbar>` / `<Words.ToolbarGroup>` no longer accept or forward `orientation`. - Six chunks of CSS that targeted `[data-orientation='vertical']` deleted (toolbar layout, content grid placement, status, find-replace, link-editor, placeholder). Demo (`web/routes/uix/components/words/+page.svelte`) - `orientations` const, `orientation` state, the chip-control UI, the `toolbarOrientation={orientation}` prop, and the corresponding line in the generated code snippet all removed. - Recipe-selector entry for `[data-orientation]` deleted. - A11y row text simplified. Verified: `npm run check` clean (only pre-existing 6 errors unrelated); soma words tests 152/152 pass. Eidos vertical CSS is gone end-to-end. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	a62b23538d	feat(words): HTML5 drag-drop block reorder + keyboard grab on grip ... The block-handle grip is now actually grabbable. Native HTML5 drag: mousedown + drag → dragstart fires → wordsDragState.sourceIndex is set → the BlockInserter switches to drop-target mode (line goes prominent, "+" hides) → dragover snaps to the nearest seam → drop commits `moveBlockTo({fromIndex, toIndex})`. Keyboard equivalent: focus the grip → Space toggles "grabbed" mode (grip lights up accent-filled, aria-pressed) → ArrowUp / ArrowDown swaps with the neighbour (via the existing moveBlock command, so each press is committed individually) → Space or Escape exits. Architecture: - `words-drag-state.svelte.ts` — module-level `$state` shared between handle (writer) and inserter (reader). Lives in eidos because this is a visual interaction pattern, not a soma/morfo contract. - `moveBlockToAt(state, from, to)` in operations.ts — splice out, then splice in at the post-removal destination. Normalizes + drops caret at the moved block's new position. - `moveBlockTo` case in commands.ts. The existing `moveBlock` (swap-with-neighbor) stays because the click-menu uses it. - BlockInserter dragover/drop handlers reuse `findSeamForCursor` — same seam math as the regular insert hover, so the drop indicator lands exactly where the "+" would have. `seam.insertIndex > fromIndex ? -1` correction accounts for the source's removal shifting subsequent indices down by one. CSS: - `[data-words-block-inserter][data-mode='drop']` makes the line solid + thicker + with an outer glow during drag, and hides the "+" button entirely (the line IS the drop indicator). - `[data-words-block-handle][data-grabbed]` paints the grip accent- filled with a focus ring for keyboard-grab state. aria: `aria-pressed={keyboardGrabbed}`, descriptive aria-label on the grip mentioning Space + arrows. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	305fd70129	fix(words): eliminate flicker on block inserter "+" button ... The previous pass listened to mousemove ON the content element and cleared the seam on mouseleave. The "+" button lives at left:-28px of the content (in the gutter), so the moment the cursor reached the button it triggered mouseleave on content → seam cleared → button vanished → cursor was over content again → mousemove → button shown → loop. That was the constant flicker. Three fixes, smallest to largest: 1. **Listen on document, not content.** A single document-level mousemove handler with manual bounds checking. No more mouseleave/mouseout firing as the cursor crosses element edges. 2. **Hold steady when over the inserter itself.** If `e.target` is under `[data-words-block-inserter]`, return early — don't recompute or clear the seam. The button can stay reachable forever. 3. **Memoize the seam.** Reuse the existing Seam object whenever the snap target (`insertIndex`) is unchanged. Cursor jitter within the same snap window no longer triggers re-renders. Bounds-check buffer: 48px to the left (so the "+" button at -28 is safely inside the keep-alive zone), 16px on the right, 12px top/bottom. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	9b1a964c69	feat(words): "+ between blocks" hover-zone inserter (POLISH-2b) ... Fixes the wedge case the user just hit: two atomic blocks back-to-back (image, table, code) leave no caret position between them, so there's no way to insert a paragraph by clicking. Now: hover near the seam between any two top-level blocks and a "+" appears on the left margin with a faint accent line spanning the seam. Click "+" → a new paragraph is dropped at that index and the caret lands at the start so the user starts typing immediately. Same overlay pattern as the block-handle: single component watches mousemove inside `[data-words-content]`, computes the boundaries of all top-level blocks (plus "before first" and "after last" seams), snaps to the closest seam within 18px of the cursor's Y, and renders a fixed-positioned strip with the "+" button and accent line. New engine surface: - `insertBlockAt(state, blockIndex, block)` in `operations.ts` — splices the block in, normalizes, drops caret at `[blockIndex, 0]`. `blockIndex === children.length` appends to the tail. - `insertBlock` case added to `WordsCommand` + dispatcher in `commands.ts`. The command surface keeps `block` as a loose JSON record so consumers don't need to import `WordsBlock` — the reducer casts on the boundary and `normalizeDocument` validates. Verified in browser: hover at the seam between code (index 3) and image (index 4) → "+" appears at the left margin → click → doc becomes 0:heading | 1:paragraph | 2:quote | 3:code | 4:paragraph | 5:image; caret in the new paragraph ready for input. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	e6caf308b0	refactor(words): refine drawer Editable — hover pencil + textarea + icon-only controls ... User feedback on the previous DRAWER-EDITABLE pass: the "Edit / Save / Cancel" labeled buttons were too heavy. The reference design is the Notion / Linear inline-edit pattern — resting state shows only the preview text, a small pencil reveals on hover, click opens a textarea with compact icon-only ✓ / ✕ controls beneath. Drawer refactor: - `activationMode="none"` + `submitMode="none"` — only the explicit triggers drive state. Removes accidental commits on blur / Enter (Enter must be a newline in a textarea anyway). - `Editable.Input` swapped to a textarea via the `child` snippet: `{#snippet child({ props })}<textarea {...props} rows={2}></textarea>`. ALT and CAPTION can run long; a single-line input was wrong. - `EditTrigger` / `SubmitTrigger` / `CancelTrigger` now hold only an icon (Pencil / Check / X) with `decorative` flag + `aria-label` on the trigger button for screen readers. - Wrapper attribute `data-words-drawer-editable` scopes all the custom CSS to these two Editables — other Editable instances in the app keep the canonical wrapper look. CSS (`words.css`, scoped to `[data-editable][data-words-drawer-editable]`): - Pencil hidden by default (opacity 0); revealed on `:hover` / `:focus-within` of the editable row. - Edit mode flips layout to column: textarea fills the row, the Control bar sits beneath, right-aligned. - Textarea has a real border + focus ring (was inheriting the ghost variant's transparent treatment, looked invisible against the drawer). - Submit chip is filled accent (✓ stands out), Cancel chip is outlined neutral. Both 28×28 square. - Preview placeholder ("Describe the image…") is italic + muted. Verified in browser: - Hover over ALT row → pencil appears at the right (opacity 0.7) - Click pencil → textarea visible with current value preselected - Below: purple ✓ + outlined ✕ - Type new value → click ✓ → `<img alt="...">` in the editor and the drawer preview both update; mode returns to preview. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	ffe2c56f42	feat(words): use Editable for ALT/CAPTION in image drawer panel ... User-suggested polish: replace the read-only `<p>` previews for the image alt and caption fields with the canonical Editable component. Same inline-edit shape used in the dedicated Editable demo — preview text with a small "Edit" trigger that swaps to an input + Save/Cancel on activation. Closer to the Notion/Figma/Linear interaction the user expects out of a sidebar property panel. Wiring: - `imageBlockAtCaret` now returns `{ index, block }` so the panel can dispatch `applyCommand({ type:'updateBlock', blockIndex, patch })` with the right index instead of re-walking the selection. - ALT: `onValueCommit={(next) => s.applyCommand({ type:'updateBlock', blockIndex, patch: { alt: next } })}`. `placeholder` "Describe the image…" surfaces when alt is empty. - CAPTION: same shape; `patch: { caption: next || undefined }` so clearing the field drops the field entirely from the block. - `size="sm" variant="ghost" color="primary"` chosen to fit the drawer's compact density without competing with the image preview. - Status (pending/error) stays as the existing readonly tag — not user-editable. Also: add `selectedBlockIndex: number | undefined` to `WordsProviderSnippetProps`. The provider was already emitting it (consumed by `imageBlockAtCaret` and the image float bar) but the type definition didn't list it. Verified in browser: clicking the image enters image mode in the drawer; clicking ALT's Edit trigger reveals the input; typing a new value and clicking Save propagated end-to-end — `<img alt="…">` in the editor and the preview both updated. Same flow for caption. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	66225404e1	feat(words): drag handle + block action menu per block (POLISH-2a) ... Notion-style :: grip rendered to the left of the hovered block via a single overlay component (`Words.BlockHandle`). Click reveals a small popover with Move up / Move down / Duplicate / Delete. Each fires via `snippet.applyCommand` so the discrete imperatives bypass the sema runtime (no extra event noise on top of what the operation already implies for downstream consumers). Backed by three new generic operations on the engine: - `moveBlockAt(state, blockIndex, direction)` — swap with neighbour. - `duplicateBlockAt(state, blockIndex)` — JSON-clone insert-after. - `deleteBlockAt(state, blockIndex)` — already existed; now wired to the menu's Delete entry. The three operations are exposed as `moveBlock` / `duplicateBlock` / `deleteBlock` commands in `WordsCommand` so any future block-toolbar can reuse them without per-feature engine plumbing. Also: add `applyCommand` / `selectAtomicBlock` / `clearSelectedBlock` to `WordsProviderSnippetProps` — they were emitted by the provider but missing from the type. Float bar + block handle both consume `snippet.applyCommand`. Verified in browser: hover over a block renders the grip (24×24, opacity 0.45 → 1 on hover, z-index popover); click opens menu with all 4 actions correctly labelled; Duplicate on the heading at index 0 produced a new heading at index 1 and pushed the rest down. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	5acee94777	fix(words): color picker uses applyCommand (no sema) + onchange instead of oninput ... Two related fixes to the color picker UX: 1. Swatches now call applyCommand({type:'toggleMark', mark:'color:#x'}) directly instead of runCommand('color:#x'). runCommand emits the 'commit-set-format' sema event (which fires sound + visual feedback on the editor surface). For discrete color picks the sema event is noise — applyCommand bypasses the trigger and just mutates the doc. 2. Native <input type=color> uses `onchange` instead of `oninput`. `oninput` fires continuously while the user drags the OS color picker — each fire was queueing a separate command + sema event, producing a horror cascade of sound / visual flashes on every pixel of the slider drag. `onchange` only fires when the user releases / commits the picker (closes the OS popup), so we get exactly one command per intent. Same pattern applied to both Foreground and Background pickers + the clear (×) buttons. Sema events are still emitted by the rest of the drawer's action chips (run via runCommand) because those are user-facing intent actions (Bold, Italic, etc.) where the perceptual feedback aligns with the user's notion of "I just did a thing". Color slider drags are NOT in that category — they're continuous parameter tuning. Verified: clicking the red swatch on selected "ActiveUI" produces the expected red <span> in the document AND the trace shows no new commit-set-format event (was previously firing on every swatch click, drowning audio + animation). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	5c10ab74b3	feat(words): floating image control bar + generic updateBlock/deleteBlock commands (POLISH-1a) ... Quick-action toolbar that hovers above the selected image, mirroring the pattern in Notion / Linear / Figma. Lets the user align L/C/R, delete, or dismiss the image without having to expand the Image panel of the drawer. New eidos component (spike — no morfo/soma yet, refactor canon if it sticks): src/uix/eidos/components/words/words-image-float-bar.svelte - Position: fixed, calculated from the selected figure's getBoundingClientRect on every selection / scroll / resize / figure size change (ResizeObserver + window listeners). - Visible only when snippet.selectedBlockIndex points to an image block; null position otherwise so the component reflows away. - 5 actions: align left / align center / align right / delete / dismiss. Divider between layout and destructive actions. - Calls snippet.applyCommand for align / delete instead of running through runCommand string-name path (commands target an arbitrary block index, no string lookup needed). Engine — new generic commands so feature panels stop reaching for hacky JSON re-imports to mutate single blocks: - WordsCommand union gains 'updateBlock' (blockIndex + patch) and 'deleteBlock' (blockIndex). - updateBlockAt(state, idx, patch) — shallow-merge patch into the block, normalize doc, clamp selection. No-op when patch matches current values. - deleteBlockAt(state, idx) — remove block, fall back to empty paragraph when doc would be empty, place caret on the next block. - applyCommand exposed on provider snippetProps so any consumer (eidos panels, future block toolbars, tests) can drive arbitrary WordsCommand without needing a string-name route. CSS additions: float bar styling (position fixed, accent border, shadow), action chips with hover + active + risk + ghost variants, divider. Demo: <Words.ImageFloatBar snippet={snippetProps} /> mounted next to <Words.Drawer />. Verified visually: clicking the demo image now (a) outlines the image (COLOR-1 highlight), (b) opens the drawer in IMAGE mode, (c) shows the float bar hovering above the figure with "align center" active (default), trash + dismiss available. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	c1c3cd9404	feat(words): parametric color marks — foreground + background pickers (COLOR-3) ... Marks now support parametric values via colon syntax: - color:#hex — foreground text color - bgcolor:#hex — highlight / background color Existing boolean marks (bold/italic/underline/strike/code) are unchanged. Empty value (color:, bgcolor:) clears that slot. Engine - WordsMark union extended with `color:\${string}` and `bgcolor:\${string}` template literal variants. - isWordsMark accepts the parametric form (validates against #hex 3/4/6/8 digits). - normalizeMarks de-duplicates by prefix for parametric marks (only one color, one bgcolor per inline survives). - toggleMarkSet REPLACES same-prefix entries instead of toggling them; empty value strips the slot. - New helpers markValue() and markPrefix() for reading parametric data. Serializers - HTML out: wrapMarks collects color/bgcolor into a single `<span style="color:..;background-color:..">` instead of nested spans, then wraps with the boolean-mark tags. - HTML in: not extended this pass — round-trip via style attr is a follow-up. The marks survive in the document model + render OK. - Markdown out: silently drops parametric marks (no MD syntax for inline color; HTML round-trip is the supported path). DOM render - createWordsTextAttrs splits marks into booleans (→ data-words-marks) and parametrics (→ inline `style` attr) so the contenteditable paints the colors directly. Drawer Format panel - New "Foreground" + "Background" sections with: - 10-swatch palette (red/orange/amber/green/cyan/blue/violet/pink/ white/grey) — hand-picked editorial set - Native <input type=color> for free hex pick (no preset constraint) - "×" clear button (only when a value is set for that slot) - Active swatch outline when the current value matches the preset - runCommand passes the `color:#hex` / `bgcolor:#hex` string straight through commandFromName → toggleMark. WordsCommandName extended with `color:\${string}` and `bgcolor:\${string}` for type-safe consumer calls. 152/152 tests pass. Verified visually: clicking the red swatch on selected "ActiveUI" text in the demo produces `<span data-words-marks="bold" style="color:#ef4444">ActiveUI</span>` in the rendered HTML, text paints red in the editor, the swatch shows active outline + clear (×) button appears. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	ab06562abe	style(words): drawer per-scope accent color + tinted header (COLOR-2) ... Each drawer panel category gets a distinct hue so the user reads the panel's *kind* at a glance: - Text scopes (format, block) → primary accent (project violet) - Structure scopes (cell, row, table, list, list-item) → secondary (slate-blue) — the contextual scaffolding around the text - Media scopes (image, code) → fulfill (teal-mint) — opaque blocks - Default (document outline) → neutral toolbar tone Implementation: scope hue resolves into `--_drawer-scope-accent` via `[data-words-drawer-panel][data-mode=...]` selectors. Then: - The left border (already used as depth indicator) inherits the scope accent at decreasing alpha as depth grows (leaf 100%, depth 1 65%, 2 40%, 3 25%). - Open panel headers paint with a 6% tint of the scope accent + the accent color for the title text. - Active action chips inside a panel inherit the scope accent for background + border + text, so the active variant matches the scope's color identity (no more "always-violet" Striped/Plain). Visual cohesion: a user in a table cell with text selected now sees: - FORMAT panel header in violet (text scope) - CELL/ROW/TABLE headers in slate-blue when expanded (structure scope) - IMAGE/CODE panels in teal-mint (media scope) - DEFAULT in neutral CSS-only change; nothing in soma/morfo touched. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	b3524b3446	fix(words): image block selection — click figure activates image drawer mode (COLOR-1) ... Image blocks are contenteditable=false so the text caret can never live inside them. The drawer never entered image mode and the user had no visible "selected image" state. Fixed with a separate atomic- block selection layer: - New provider state `selectedBlockIndex` (number | undefined) for atomic block selection. Distinct from text selection. - `currentBlock` derived now short-circuits to the atomic block's type when `selectedBlockIndex` is set, so the drawer enters image mode whenever the user has selected an image. - New methods `selectAtomicBlock(blockIndex)` + `clearSelectedBlock()` on the provider. selectAtomicBlock also focuses the editor root so keyboard handlers (esc, delete) work on the atomic block. - `syncSelectionFromDom` clears the atomic-block highlight only when the text selection actually moved (was clearing on every selectionchange echo, which wiped the highlight a tick after the click landed). - Provider onclick handler detects clicks on figure[data-words-block= 'image'][data-words-path], decodes the path, calls selectAtomicBlock(blockIndex). - Render passes `selectedBlockIndex` through findOpts; image block render adds `data-words-block-selected` attr when the index matches. - Eidos CSS paints a 3px accent outline + glow on the selected figure. Image panel's imageBlockAtCaret() now falls back to selectedBlockIndex first, then text-selection anchor. Result: clicking an image in the editor now highlights it (accent outline + glow), the drawer enters IMAGE mode and shows its source/alt/caption/status panel. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	e25ac301b5	feat(words): drawer hierarchical mode stack — Format → Cell → Row → Table (HIER 1-3) ... The drawer now renders the FULL hierarchy of scopes that apply at the caret, instead of only the most-specific one. Inside a table cell with text selected the user sees four panels stacked: Format ← active selection's inline marks/links Cell ← tone, vertical align, cell text-align Row ← +/- row, row header toggle Table ← striped, compact, +/- column, col header Each panel is independently collapsible. Default: the leaf (most specific) is open, the rest are collapsed but accessible via the chevron next to their header. User-explicit toggles persist across context changes. HIER-1 — soma: - WordsDrawerMode union expanded: split 'table' → 'cell' | 'row' | 'table'; split 'list' → 'list-item' | 'list'; added 'block' for the paragraph/heading/quote text-align scope. - WordsDrawerProvider gains `modes: readonly DrawerMode[]` derived that walks the caret context outward (most-specific first). The legacy `mode` field is kept as the top of the stack for consumers that just want the leaf. - snippetProps surfaces both `modes` (stack) and `mode` (leaf). - Morfo `data-words-drawer-mode` values updated to the new union. HIER-2 — eidos panel split: - Table panel decomposed into three: cell-scope (tone + vertical + text-align), row-scope (+/- row, row header), table-scope (+/- col, col header, striped, compact). - List panel decomposed into list-item-scope (indent/outdent) + list-scope (kind). - New 'block' panel hosts text-align when caret is in a heading / paragraph / quote outside a table cell. - Each panel is a `<section data-words-drawer-panel data-depth=N>` with its own toggleable header + body snippet. HIER-3 — per-panel collapse: - SvelteMap stores user-explicit open/closed overrides per mode. isPanelOpen() falls back to "open if leaf, else collapsed" when no override exists. - Drawer header now shows leaf title in bold + breadcrumb of deeper panels in muted text (Format · Cell · Row · Table). - Visual depth cue: border-inline-start of each panel renders at decreasing strength as depth grows (leaf solid, deeper at 60% / 30% accent border). Visual verification: selected text inside a sample table cell shows 4 panels stacked (Format open, Cell/Row/Table collapsed), header breadcrumb "FORMAT · Cell · Row · Table". Click on any collapsed header expands it. 214/214 tests pass. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	9a5f3c2359	refactor(words): drawer canon — morfo part + soma provider + eidos consumes via provider (DRAWER-CANON) ... Closes the architectural gap from the DRAWER spike: the companion drawer now follows the same morfo/soma/eidos pattern as every other Words part (Toolbar/BubbleMenu/SlashMenu/LinkEditor/Status/etc.). DRAWER-CANON-1 — morfo + soma: - wordsMorfo gains a 'Drawer' part (kebab 'drawer', archetype 'group', defaultElement 'aside', role 'complementary'). Declares data-words- drawer-mode (default/format/image/table/code/list), data-open, data-disabled, aria-label. New event 'commit-toggle-drawer' fires on collapse/expand for sema/telemetry hooks. - New WordsDrawerProvider in words-provider.svelte: owns `open` bindable state, derives `mode` from parent provider's currentBlock + selection (same logic as the spike, now in soma where it belongs), exposes snippetProps {mode, open, disabled, snippet, toggle, setOpen}. `toggle()` and `setOpen()` emit the event. - WordsDrawerProps + WordsDrawerSnippetProps + WordsDrawerMode types added to types.ts and re-exported via exports.ts. - soma/components/words/components/words-drawer.svelte (the soma component) mounts the provider and forwards merged props to either the `child` snippet or the default <aside> render. DRAWER-CANON-2 — eidos refactor: - eidos words-drawer.svelte rewritten as a consumer of <Words.Drawer>. The shell (aside + header + collapse button) and the per-mode body now live inside the children snippet that receives the soma's drawer snippet props. Mode derivation removed from eidos (it's now in soma). Local open $state removed (now bindable via the soma component). - Demo +page.svelte: `<Words.Drawer snippet={snippetProps} />` collapses to plain `<Words.Drawer />` — the soma provider hands the parent snippet down internally. DRAWER-CANON-3 — verify: - 214/214 tests pass in soma+morfo scope (no regressions). - npx tsc --noEmit clean for words area. - Visual verification: drawer still renders with DOCUMENT/OUTLINE/ STATS/SUGGESTIONS in default mode; context-switching to table / code / list / format still works; aside element now carries role='complementary', aria-label from morfo translationRef, data-words-drawer-mode reflecting current panel. The eidos drawer now passes the same canon checks as every other Words part: morfo declares its contract, soma owns the state, eidos only paints. The schema validator can verify the drawer's attrs, the eidos lint can check selector usage, sema can react to the toggle event. Demo header chips reflect the new contract: PARTS 19→20, EVENTS 11→12. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	15dbf0f8fd	feat(words): drawer contextual panels — format/image/table/code/list (DRAWER-2) ... The companion drawer now switches its body content based on where the caret is. Five context modes wired to runCommand on the snippet provider, all reading current state to show active highlights: - format: B/I/U/S/code mark toggles (data-active reflects activeMarks), align L/C/R/J (data-active reflects currentTextAlign), Add link (URL prompt) + Remove (canUnlink-gated), Clear formatting. - image: read-only view of src, alt (italic + "(missing)" placeholder when empty), caption, status (pending → accent tint, error → threat tint). Edit-image flow deferred to a follow-up (needs a new updateImage operation; for the spike, view-only is enough to validate the panel works). - table: Rows +/- · Columns +/- · Row/Col header toggles · Striped / Compact toggles · Cell tone Plain/Muted/Accent. Every toggle reads current* state so the user sees which buttons are currently applied. - code: Plain + TS/JS/CSS/HTML/JSON/Bash/Py language chips, data-active reflects currentCodeLanguage. - list: Bullet/Numbered/Check kind chips · ⇤ Outdent / ⇥ Indent. All buttons use onmousedown preventDefault so the editor selection isn't dropped when clicking a drawer action (would otherwise lose the target paragraph / table cell / list item). CSS adds a generic .data-words-drawer-action chip (with hover, active, ghost, full, risk tones) + readonly-text style with pending/error tints. Reused across all panels, so adding new panels later means markup-only, no CSS extension. Verified in browser: - Click table cell → drawer shows TABLE panel with Striped active (matches sample's striped=true) and Plain tone active (matches default cell tone). - Click code block → drawer shows CODE panel with TS chip active (matches sample's language='ts'). - Click list item → drawer shows LIST panel with kind + indent actions. Concept A+D validated visually: ONE surface replaces what would otherwise be bubble menu + slash menu + heading picker + language picker + table popover + image controls + link editor. The contextual switch happens automatically; the user just moves the caret. DRAWER-3 (polish): mobile bottom-sheet layout, A/B toggle in demo controls, image-edit flow — deferred to next session if needed. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	23f90a775c	feat(words): companion drawer right-rail (concept A+D) — shell + default panel (DRAWER-1) ... A radical disposition shift: instead of N fragmented popovers (bubble menu + slash menu + link editor + heading picker + language picker + table popover + image controls) we mount ONE contextual right-rail that changes its panel based on what the user is editing. Spike-1 surface (this commit) ships: - New eidos component <Words.Drawer> (words-drawer.svelte) — a collapsible right sidebar with mode detection derived from snippet props (currentBlock + selection). 6 modes mapped: default | format | image | table | code | list. - Default panel (visible when no contextual mode active): - OUTLINE — flat list of h1/h2/h3 headings, indented by level - STATS — word count + char count + reading time (200wpm) - SUGGESTIONS — writer-coaching warnings derived from the document model: missing alt text, long paragraphs (>600 chars), heading skip patterns (h3 without h2 above). Empty state shows "Nothing flagged" in affirm color. - Contextual panels (format/image/table/code/list) render a placeholder "coming soon" message — implemented in DRAWER-2. - Eidos CSS: - Provider host gains a grid-template-areas layout when a drawer is present: 'toolbar drawer / content drawer / status drawer'. Existing children (toolbar/content/placeholder/status) get explicit grid-area assignments. - Drawer collapses to 2.25rem (just the chevron) when closed. 16rem default open. - Outline items, stats grid, warning list each have their own discrete styling. Demo (+page.svelte): - New `showDrawer` state (default true). - Children snippet refactored to capture full snippetProps via @const destructure (kept selectedLink working). - <Words.Drawer snippet={snippetProps} /> mounted inside the children block alongside the other parts. Visual verification: - Drawer renders as a dedicated right column with the editor content on the left, no layout overlap. - DOCUMENT header + OUTLINE shows "H1 ActiveUIX Words", STATS shows 73 words / 508 chars / 1 min reading, SUGGESTIONS shows "Nothing flagged" (sample doc has no warnings). - Collapse button toggles open/closed via animated inline-size transition. Next: - DRAWER-2: format/image/table/code/list contextual panels wired to runCommand. - DRAWER-3: polish (mobile drawer-as-bottom-sheet, transitions, A/B toggle in demo to compare classic toolbar vs companion drawer). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	ef95b1a82a	feat(words): slash command /image + paste/drop with onUploadImage (F3.7, F3.8) ... F3.7 — slash command /image: - DEFAULT_SLASH_COMMANDS gains an 'image' entry: triggers a window prompt for the URL (and optional alt text), then dispatches insertImage. The native prompt is MVP — consumers wanting a custom dialog can hide the slash entry via slashCommands override and call applyWordsCommand({ type: 'insertImage', src }) from their own UI. - WordsSlashCommandId union + WordsCommandName union extended with 'image' / 'insert-image' respectively. - slashCommandToWordsCommand returns undefined for 'image' so the prompt path in commitSlashCommand owns it. F3.8 — paste/drop with onUploadImage callback: - New WordsProps.onUploadImage prop: (file: File) => Promise<{url, alt?}>. Consumer-injected upload pipeline (S3, R2, own backend, etc.). - onpaste detects image files in clipboardData.files. If present + the callback is wired, intercepts the paste (drops text/html processing) and inserts each image with a blob: URL + status='pending'. - New ondragover + ondrop handlers gate on dataTransferHasImageFiles and the callback being set; trigger insertImageFiles on drop. - insertImageFiles inserts pending placeholders, awaits the upload promise for each, then either swaps src for the final URL (success) or flips status='error' (rejection). Matches blocks by blob URL (unique) so concurrent uploads + concurrent edits stay coherent. - replaceImageBlock / markImageBlockError use mapImageBlocksBySrc to rewrite the document via replaceDocument command. No new command type needed. - Helpers added: collectImageFiles, dataTransferHasImageFiles, mapImageBlocksBySrc — all pure functions at module scope. Test fixture: wordsOpts gains the onUploadImage state slot so the 16 existing provider tests typecheck without behavior change. 152/152 tests pass in the full words soma scope. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	86b47cef25	fix(words): normalize handles image blocks + 3 image tests (F3.11 partial) ... normalizeBlock had no case for type='image' — image blocks coming from parseWordsMarkdown / parseWordsHtml were silently dropped at the default branch. Now: - normalize.ts case 'image': validates src (non-empty + sanitizeWordsUrl), passes through alt / caption / width (numeric) / height (numeric) / align (validated via isWordsImageAlign) / status ('pending' | 'error'). - normalize.ts imports createImage + isWordsImageAlign from the extension. Image-extension cast widened with `as unknown as ErasedFactory` because WordsImageBlock (no `children` field) doesn't satisfy the contract's node shape strictly; the double-cast is the same pattern table-extension already uses for typed factories. 3 new engine tests cover the F3 happy path: - parses a block-level markdown image ( on its own line yields a type:'image' block) - round-trips an image through markdown (serialize → parse → matches) - round-trips an image through HTML (figure>img+figcaption → ImageBlock → serialize back) 85/85 engine tests pass (was 82; +3 image tests). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	2a51adc00f	feat(words): insertImage operation + command + CSS + extension stub (F3.6, F3.9, F3.10) ... F3.6 — insertImage reducer + command: - engine/operations.ts: insertImage(state, options) — atomic insert (URL required, optional alt/caption/width/height/align/status). Replaces an empty paragraph cursor lands on, otherwise inserts after. Caret lands on the new paragraph following the image (image is atomic — no caret inside the figure). - engine/commands.ts: 'insertImage' case added to WordsCommand union + applyWordsCommand switch. Public command surface: provider can call applyWordsCommand(state, { type: 'insertImage', src, alt?, ... }). F3.9 — eidos CSS for image block: - figure[data-words-block='image']: centered flex column with border, bg, padding, gap. - img: max-inline-size 100%, auto height, user-select none. - figcaption: italic, centered, status-color, smaller font. - data-words-image-align='left'/'right': flex alignment + auto margins. - data-words-image-status='pending': opacity + grayscale + "Uploading…" ::after pseudo-text. - data-words-image-status='error': threat border + "Upload failed" ::after pseudo-text. - :focus / :focus-within: 2px accent outline so the atomic block has a visible selection state. F3.10 — imageExtension stub: - extensions/image/image-extension.ts: WordsExtension shape with name='image', nodeTypes=['image'], factories.image=createImage, commandNames.insertImage='insertImage'. Hooks intentionally absent (engine still hardcodes branches today; stub publishes the public shape parallel to tableExtension). - extensions/index.ts exports imageExtension alongside tableExtension. Morfo: - data-block + data-words-node value sets include 'image'. - New optional attrs declared: data-words-image-align (left/center/ right), data-words-image-status (pending/error). Now the eidos CSS selectors validate against the morfo contract. 211/211 tests pass in words soma + morfo scope. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	08bef77f4b	feat(words): image extension foundation — types, render, HTML+Markdown serializers (F3.1-F3.5) ... Adds image as a first-class block node to the words engine. Phase F3 of the words rich-text editor. Ships the bottom half of the stack so images can be authored in source (markdown / HTML / value JSON) and roundtripped, but UX surfaces (toolbar action, slash command, paste/ drop) land in subsequent commits. F3.1 — extensions/image/{types,factories,index}.ts: - WordsImageBlock = { type: 'image', src, alt?, caption?, width?, height?, align?, status? } - createImage(src, options) factory canonicalising defaults (align='center' stripped, etc.) - WordsImageStatus = 'pending' | 'error' for upload lifecycle - WORDS_IMAGE_ALIGNS + isWordsImageAlign predicate F3.2 — engine/document.ts: - WordsBlockType + WordsBlock union extended with 'image' - Image types re-exported (backward-compat with the table-types pattern) - getBlockText: image → alt text (captions excluded; image stays a single atom in plain-text contexts) F3.3 — engine/render.ts: - Image branch renders <figure data-words-block=image contenteditable=false> <img src alt width? height? loading=lazy draggable=false> <figcaption>caption</figcaption? - WordsRenderTag union gains 'figure' | 'img' | 'figcaption' - data-words-image-align / data-words-image-status surface state - renderBlockPlainText: image → alt F3.4 — engine/serialize-html.ts: - Out: <figure data-words-image-align?><img alt src width? height?><figcaption?></figure> - In: 'figure' added to BLOCK_TAGS; <figure><img> elements parsed back to ImageBlock (with caption from <figcaption>); bare <img> parses to a block too - htmlElementText helper added for caption text extraction F3.5 — engine/serialize-markdown.ts: - Out:  — escapes ], ( and ) in src, " in caption - In: standalone line matching ^$ promotes to a block image. Inline images mid-paragraph become text + no image (lossy by design — keep the model tight; inline image is a separate node type if added later) Engine ripple — exhaustive switches updated: - normalize.getInlineBlockText - path.collectContainerChildren (image marked as terminal) - selection.collectBlockText (image returns empty — opaque atom) - operations: - setBlock skips image (image is not convertible to inline-text blocks) - setTextAlign skips image (no text align on image) - deleteRange refuses if range crosses an image (atomic) - insertParagraph on image inserts a fresh paragraph after it and moves the caret - blockTextAlign() returns undefined for image - words-provider.currentTextAlign returns 'left' on image 149/149 tests pass in the full words soma scope. `npx tsc --noEmit` clean for the touched area. Engine consumers compile unchanged. Next: F3.6 insertImage operation + toolbar/slash command, F3.7 slash menu /image entry, F3.8 paste/drop with onUploadImage callback, F3.9 eidos CSS, F3.10 imageExtension stub, F3.11 demo + verify. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	5bd9f9fe4f	fix(words): slash menu wider + description wraps to 2 lines instead of ellipsis (F4.5a) ... Before: - Menu width was 7× button-size (≈224px), so even short descriptions clipped: "Large section headin…", "Blockquote callou…". - Description was single-line nowrap+ellipsis, doubling the clipping. After: - Menu width 10× button-size (≈320px) — enough room for the longest default description ("Large section heading") without truncation. - Description switches to a 2-line line-clamp with line-height 1.3, so longer custom descriptions still degrade gracefully instead of silently disappearing past the first line. Label still nowrap+ellipsis (the bold first line is supposed to be a short title; if it doesn't fit, that's an author issue worth a visual cue). CSS-only in words.css. Visible immediately on the demo's SlashMenu. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	eb6a019b5a	fix(words): vertical rhythm — heading top margin, blockquote distinctive, code block generous padding (F4.3) ... Three small typography lifts that compound into a noticeably cleaner document rhythm: - Headings (h1/h2/h3) gain 1.5× block-gap top margin when NOT first child. Opens visible breathing room between a body block and the next section. Skipped on first-child so the editor opens flush against the toolbar. - Blockquote: was just a left border + recolored text. Now adds vertical padding, italic, a 1.5× border-inline-start width, and a 4% accent background tint. Reads unambiguously as a callout rather than indistinguishable from a regular paragraph. - Pre (code block): padding bumped from 1× quote-px to 1.25× and the border mixed with 60% accent-border so the block is visually weightier and stands apart from the body bg. CSS-only change in words.css. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	c09f7b0d87	fix(words): table accent tone softer, hover cells, stronger header divider (F4.2) ... Before: - tone='accent' painted at 14% solid + full accent-text recolor — the cell read as "broken / special-cased random" rather than emphasised. - striped used 70%-transparent overlay on toolbar-bg ≈ invisible against the editor bg. - no hover state on cells — users had no signal that cells are interactive (clickable to edit). - header row used the same single-width border as body cells, so the divider got lost when striped or accent kicked in. After: - tone='accent' at 8% solid, text color inherited from default — the cell still draws the eye without shouting. - striped uses solid toolbar-bg (no transparency) — actual zebra. - subtle hover (5% accent overlay) on every non-header cell, with a 120ms ease transition. - header row gets a 2× bottom border so the head/body divider is unambiguous regardless of striping/accent. CSS-only change in words.css, no markup or logic touched. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	3cc171f747	fix(words): toolbar horizontal-scroll instead of multi-line wrap on mobile (F4.1) ... Before: at narrow viewports the toolbar wrapped to 2-3 lines with orphan icons in the last row (e.g. at 375px: 5 + 4 + 1 icons = 3 rows). Awkward visually and wasteful of vertical space. After: single-row horizontal scroll with a soft trailing fade (mask-image gradient) hinting at hidden overflow. Same UX as Google Docs / Word / Notion mobile toolbars. All commands stay reachable without changing the markup or the provider logic. Vertical-orientation toolbar opts out of the horizontal-scroll mask (it already lays out top-to-bottom and the mask would clip the trailing icons in the visible direction). Verified: - 375px (mobile) — single row, fade visible on right edge. - 900px (laptop) — single row, no overflow, fade harmless. - Vertical orientation — column layout intact. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	1043f1de4a	refactor(words): extract 8 pure table helpers to extensions/table/utils.ts (F2.3h.1) ... Moves the table-pure helpers (no engine selection/normalize/marks dependencies) from `engine/operations.ts` to `extensions/table/utils.ts`: - createTableGrid (2×2 default, clamped 1..12 × 1..8) - tableCellPathFromContainer (3-tuple path resolver) - tableOptions / tableCellOptions (option-bag readers) - tableColumnCount (max-cells-across-rows, min 1) - createEmptyTableRow (N-cell row factory) - insertTableCellInRow / deleteTableCellFromRow (cell ops over a row) - WordsTableCellPath type alias The 6 *mixed* helpers (currentTableCellPath, selectTableCell, setTableCellTextAlign, updateCurrentTableCell, updateCurrentTable, preserveTableFormattingSelection) still live in operations.ts — they read private engine state and can only move once F2.3i full lands a visitor pattern that injects those engine APIs as parameters. clampInteger stays duplicated locally in utils.ts (9 LoC, generic numeric helper) instead of being exported from the engine, to keep the extension's surface tight. If a third caller emerges, promote to a shared location. 149/149 tests pass in the full words soma scope. TS check clean in extensions+engine area. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	bde6728590	refactor(words): table extension stub — name + nodeTypes + factories + commandNames (F2.3i) ... Publishes `tableExtension: WordsExtension` in `extensions/table/table-extension.ts` with the three static facets we can ship without touching the engine: - `nodeTypes`: ['table', 'table-row', 'table-cell'] - `factories`: typed constructors for the three node types - `commandNames`: the 12 public table commands (insertTable, deleteTableRow, toggleTableHeaderRow, setTableCellTone, …) mapped to internal operation types (1:1 today). The dynamic hooks (`render`, `normalize`, `serialize`, `path`, `keyboard`, `commands`, `events`) are intentionally absent — F2.3f-h port each one incrementally. The engine does NOT yet consume the registry; the stub publishes the shape the engine will respect once F2.3i full lands. 6 smoke tests verify the public shape (nodeTypes, factories produce valid nodes, registry registers cleanly + finds by every nodeType, all 12 command names exposed, hooks intentionally absent, default 1×1 table). 102/102 in extensions + engine scope, 149/149 in full words soma scope. CONTINUE.md updated to reflect F2.3i-stub done + next-session order (F2.3h first, then F2.3f/g — F2.3i full wire-up follows F2.3h). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	d839273536	refactor(words): extract table render to extensions/table (F2.3e) ... Fifth atomic checkpoint of F2.3. Moves the table-specific DOM render helpers from engine/render.ts to extensions/table/render.ts. Moved: - renderTable(block, path, helpers) — replaces the table branch in renderWordsBlock + internal renderWordsTableRow + renderWordsTableCell - renderTablePlainText(block, helpers) — replaces the table branch in renderBlockPlainText The extension defines structural render-element types (TableRenderElement, TableRenderText, TableRenderNode) assignment- compatible with the engine's WordsRenderElement/Node. The engine passes its constants (WORDS_PATH_ATTR, WORDS_NODE_ATTR) + helpers (encodePath, renderInlines, getInlineText) via a TableRenderHelpers object, keeping the extension independent of engine/dom.ts. The engine's renderWordsBlock dispatcher branch for `block.type === 'table'` now delegates to `renderTable(block, path, { ... })` in a single call. ~50 LoC of helper functions removed from the engine. Extension barrel updated to expose: - renderTable / renderTablePlainText - TableRenderElement / TableRenderNode / TableRenderText / TableRenderHelpers / RenderPath types Verification: - 143/143 tests pass in src/uix/soma/components/words - 0 new TypeScript errors F2.3 progress: a + b + c + d + e done. Remaining: f (path), g (normalize), h (operations — largest at ~400-500 LoC), i (registry wire-up), j (verify). Operations remains the heaviest piece — likely its own dedicated session. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	5797e69b88	refactor(words): extract table HTML serializer (F2.3c) ... Fourth atomic checkpoint of F2.3. Moves the HTML pipe-table serialize + parse helpers from engine/serialize-html.ts to extensions/table/serialize-html.ts. Moved (write side): - serializeTableHtml(block, serializeInlines, escapeHtmlAttr) - serializeTableCellHtml (internal) - tableAttrsHtml (internal) - tableCellAttrsHtml (internal) Moved (parse side): - parseTableHtml(node, parseInlines, textAlignFromStyle) - collectTableRows (internal) - htmlRowToWords (internal) - htmlCellToWords (internal) - tableCellVerticalAlignFromNode (internal) - tableCellToneFromNode (internal) Mirrors the F2.3d pattern: extension consumes engine inline serializer + HTML attribute escaper as callback parameters to avoid circular dependencies. The extension defines its own minimal `WordsTableHtmlElement` / `WordsTableHtmlChild` types structurally compatible with the engine's internal HtmlElementNode/HtmlNode — engine passes its own parsed nodes via a safe `as unknown as` cast at the call site (commented). engine/serialize-html.ts now contains only the dispatch call to the extension's serializeTableHtml / parseTableHtml. ~150 LoC of helper code removed from the engine; same LoC + clearer separation in extension. Extension barrel updated to expose: - serializeTableHtml / parseTableHtml - WordsTableHtmlElement / WordsTableHtmlChild types Verification: - 143/143 tests pass in src/uix/soma/components/words - 0 new TypeScript errors F2.3 progress: a + b + c + d done. Remaining: e (render), f (path), g (normalize), h (operations — largest), i (registry wire-up), j (verify). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	d6f0af4846	refactor(words): extract table markdown serializer (F2.3d) ... Third atomic checkpoint of F2.3. Moves the markdown pipe-table serialize + parse helpers from engine/serialize-markdown.ts to extensions/table/serialize-markdown.ts: Moved: - serializeTableMarkdown(block, renderInline) - tableMarkdownRow (internal) - escapeMarkdownTableCell (internal) - collectTable(lines, startIndex) - parseTableLine (internal) - isTableSeparator (internal) - NEW: buildTableFromMarkdownRows(rows, parseInline) — convenience for the engine to convert collectTable's raw output into a WordsTableBlock, keeping the extension free from the engine's inline parser implementation. The extension takes inline serialize/parse functions as callback parameters (`renderInline`, `parseInline`) instead of importing the engine's inline serializers. Avoids circular dependency — the extension imports engine for shared types (WordsInline, WordsTableRow) and factories only, never for runtime function dependencies. engine/serialize-markdown.ts now imports the 3 extension functions and calls them from `serializeBlockMarkdown` (write side) and `parseWordsMarkdown` (parse side). Internal `collectTable`, `parseTableLine`, `isTableSeparator` removed. Extension barrel updated to expose the new functions: - serializeTableMarkdown - parseTableMarkdownRows (collectTable re-exported under clearer name) - buildTableFromMarkdownRows Verification: - 143/143 tests pass in src/uix/soma/components/words - 0 new TypeScript errors F2.3 progress: a + b + d done. Pending: c (serialize-html), e (render), f (path), g (normalize), h (operations — largest), i (registry wire-up), j (verify). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	a0afa44da4	refactor(words): extract table factories to extensions/table (F2.3b) ... Second atomic checkpoint of F2.3. Moves table factory functions (createTable, createTableRow, createTableCell), predicate functions (isWordsTableCellVerticalAlign, isWordsTableCellTone), and value-set constants (WORDS_TABLE_CELL_VERTICAL_ALIGNS, WORDS_TABLE_CELL_TONES) from engine/document.ts to extensions/table/factories.ts. engine/document.ts now re-exports the moved symbols so existing imports keep working unchanged. The factories module imports its back-references (createText, WordsInline) from engine/document — the extension may depend on the engine core but the engine core only depends on the extension via re-export indirection (no functional dependency). extensions/table/index.ts barrel updated to expose types (F2.3a) + factories (F2.3b) as the table extension's public surface. Document header notes which pieces have been migrated and which are still pending (serializers, render, path, normalize, operations, registry wire-up). Verification: - 143/143 tests pass in src/uix/soma/components/words - 0 new TypeScript errors - No engine behaviour change (symbol relocation only) F2.3 progress: a + b done. Pending: c (serialize-html), d (serialize-markdown), e (render), f (path), g (normalize), h (operations — largest), i (registry wire-up), j (verify). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	088d31ec5d	refactor(words): extract table types to extensions/table/types.ts (F2.3a) ... First atomic checkpoint of F2.3 (table migration to extension). Moves the type declarations only — interfaces and unions for the table node tree (WordsTableBlock, WordsTableRow, WordsTableCell, WordsTableOptions, WordsTableCellOptions, WordsTableCellVerticalAlign, WordsTableCellTone) go from engine/document.ts to extensions/table/types.ts. engine/document.ts re-exports the moved types via `export type { ... } from '../extensions/table/types'` so every existing consumer keeps compiling without changes. Internal usages within document.ts use a separate `import type` to keep the types in local scope for factory function signatures. The rest of F2.3 (factories + serializers + render + normalize + path + operations + registry wire-up) requires deeper dependency analysis across the engine's ~990 LoC of table code spread over 10 files. That work is deferred to a dedicated session to avoid mid-refactor breakage. Verification: - 143/143 tests pass in src/uix/soma/components/words - 0 new TypeScript errors - No engine behaviour change (types-only move) Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	1805b0817d	feat(words): extension system skeleton (F2.1 + F2.2) ... Phase 1 of the words rich-text editor extension system. Skeleton + contract only — engine is NOT yet wired to consult the registry. That happens in F2.3 (migrate table) and F2.4 (migrate code-block). New module `src/uix/soma/components/words/extensions/`: - extension-types.ts: WordsExtension interface + sub-types. Every hook is optional. Hooks cover document semantics (nodeTypes, factories), normalization, commands (reducer pattern), rendering, serialization (HTML + Markdown, in + out), path navigation, keyboard, sema events, dispose. Each hook documented inline with rationale. - extension-registry.ts: createWordsExtensionRegistry() returns the runtime container. API: register / unregister / dispose extensions / findByNodeType / findByCommandName getRender / getNormalize / getCommand / getSerialize / getPath / getKeyboard allEvents / allCommandNames / allNodeTypes tryDeserializeHtml / tryDeserializeMarkdown Duplicate-name and duplicate-nodeType detection at register-time with rollback-safe semantics. dispose() tears down extensions in reverse-registration order. - extension-registry.test.ts: 14 tests covering register, conflict rejection (duplicate name, duplicate nodeType, duplicate command name), all-or-nothing rollback, sema event aggregation, dispose ordering, unregister, getRender/getNormalize/getCommand walks, tryDeserializeHtml walk-until-non-null. - index.ts: barrel. Verification: - 14/14 registry tests pass - 82/82 existing engine tests still pass (no engine touched) - Total 96/96 in src/uix/soma/components/words F2.1 (audit) + F2.2 (skeleton) done. F2.3 (migrate table) and F2.4 (migrate code-block) are separate sessions per the planned rollback-safe sub-task structure — each one will move the ~990 LoC of table and ~790 LoC of code-block from the engine core into extensions/{table,code-block}/* and wire them into the registry. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	85b8a56ab0	feat(words): add toolbarLayout grouped|inline (responsive) ... New optional prop on <Words> for the toolbar layout strategy: - 'grouped' (default): each family (text/block/list/align/insert/table/ link/tools) becomes a single trigger button that opens a popover with its commands. Compact, recommended for narrow viewports. - 'inline': every command in every family renders flat in the toolbar, with the toolbar group separators marking family boundaries. High discoverability, recommended for wide editors. Tools that own their own panel UI (LinkEditor, FindReplace, HeadingPicker, CodeLanguagePicker) remain popovers/dropdowns in both layouts by design — they are not "inlineable". Accepts ResponsiveProp<WordsToolbarLayout> for breakpoint-aware switching, e.g. `toolbarLayout={{ base: 'grouped', md: 'inline' }}`. Implementation: - types.ts: new WordsToolbarLayout = 'grouped' | 'inline'; new toolbarLayout?: ResponsiveProp<WordsToolbarLayout> on WordsProps. - index.ts: export WordsToolbarLayout type. - words.svelte: resolved via eidos.resolve(toolbarLayout, 'grouped'). New helper flattenFamilyToLeafItems(family, customItems) returns the default leaf items per family for the inline expansion. Render refactored: leaf rendering extracted to a Svelte snippet reusable by both layouts, so the conditional only chooses between <ToolbarFamilyMenu> (grouped) and {#each leafs as leaf}{@render leafItem(leaf, snippetProps)}{/each} (inline). - demo: toolbarLayout state added + wired to <Words>. Verified visually in browser (preview + screenshot): - grouped: 9 family menu triggers + 1 sibling tool = 10 toolbar buttons - inline: 0 family triggers + 28 command buttons + 5 tools = 33 buttons Tests: 161/161 pass in src/uix/{contracts,soma/components/words}. TypeScript: 0 new errors. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	58c2c0babf	fix(words): close audit P0 + quiet focus event (FASE 1) ... Cierre de los hallazgos P0/P1 del audit Codex sobre Words + fix del evento contact-focus molesto reportado por el usuario. F1.1 — Fix evento foco (UX): El handler onfocus disparaba `runtime.trigger('contact-focus')` cada vez que el editor recibía foco, generando sound + haptic + visual stamp en cada re-focus desde popovers internos (bubble menu / link editor / find panel / slash menu / toolbar dropdowns). En sesión normal de edición el foco entra/sale del content docenas de veces, lo que producía fatiga perceptual. Cambios en words-provider.svelte.ts: - onfocus: emit solo en transición real (was-not-focused → focused). Re-focus desde popover interno NO emite. - onblur: si el related target es interno al Words tool tree (isInsideWordsTool), mantener focused=true (Words se trata como un solo scope de foco). Skip commit('blur') también para evitar commits espurios entre popover hops. F1.2 — Declarar 8 data-* hardcoded en morfo (audit P0-1): Añadidos al Content part: - data-words-find-match, data-words-find-active (mark highlights) - data-words-indent (list items con indent > 0) - data-language (alias parsing-only para interop con HTML externo emitido por Prism/highlight.js) Nuevas 6 parts compound (Eidos wrappers): - HeadingPicker + HeadingPickerPanel (data-words-heading-picker / -panel) - CodeLanguagePicker + CodeLanguagePanel - ToolbarFamilyMenu + ToolbarFamilyPanel F1.3 — Rename i18n keys a kebab-case (audit P0-1): 4 keys camelCase eran ilegales según contracts.test.ts. Renamed: - components.words.bubbleMenu → bubble-menu - components.words.slashMenu → slash-menu - components.words.linkEditor → link-editor - components.words.findReplace → find-replace Propagado a: morfo declarations (8 refs), provider tslangs calls (4 refs), catalog langs/components/words.ts (4 keys). F1.4 — Canonizar 9 event names a forma {family}-{verb}[-{variant}] (audit P2-1): Renombrados para parsear con el morfo:vocabulary script. El verb declarado en cada semantic ahora aparece explícito en el name: - commit-content → commit-save-content (verb: save) - commit-format → commit-set-format (verb: set) - commit-link → commit-set-link - commit-unlink → commit-reset-link (verb: reset) - commit-slash-command → commit-set-slash - commit-check → commit-set-check - commit-history → commit-restore-history (verb: restore) - commit-clear → commit-reset-content (verb: reset) - signal-invalid-input → signal-warn-invalid (verb: warn) Propagado a: sema cascade selectors, soma provider trigger calls, provider + content tests, demo route. 0 conflictos con event names de otros componentes (file-upload/tag-group/etc. usan namespaces propios sin colisión con words). Tests: 161/161 pass en src/uix/{contracts,soma/components/words}. Suite full eidos+morfo+soma+sema pendiente de verificar tras commit. contracts.test.ts ahora verde — cierra P0 audit Codex. morfo:vocabulary 0 event-name warnings (los 107 data-attr enum warnings son pre-existentes, separate scope). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	bd73fb98f3	feat(eidos): variants canon — EIDOS_VARIANTS + per-component lint ... Variants son canon del eidos, NO del theme. Decisión arquitectónica firmemente sostenida: el vocabulario de variants (solid/outline/ghost/ soft/surface/line/pills) está fijo a nivel del framework — paralelo a las 8 sema families del libro. Theme = retintar lo perceptualmente fijo; cambia QUÉ color es `affirm`, no QUÉ significa `outline`. Cambios: - lib/types.ts: nueva constante `EIDOS_VARIANTS` con los 5 archetypes canónicos (control / selection / chip / marker / tabs). Los 5 union types se derivan via `[number]` indexed access — valor y tipo no pueden desincronizarse. Nueva `EIDOS_VARIANT_VALUES` Set flat con todos los valores canónicos + utilidades cross-component (`plain`, `subtle`). - recipe-css-contract.test.ts: nuevo test "variant CSS selectors per component match the declared type union". Por cada componente: extrae el union type de `components/{c}/types.ts` (soporta literal unions + archetype aliases; cae a advisory mode en Extract<> y conditional types); compara con `[data-{c}][data-variant='X']` selectores en `{c}.css`; reporta typos y unauthorized extensions bidireccionalmente. - THEMING.md §19: nueva sección "Variants son canon del eidos, NO del theme" con argumentación (portabilidad, type safety, archetypes perceptuales paralelos a sema families), tabla de las 3 capas de la cebolla, referencia a `EIDOS_VARIANTS`, comparación con Radix Themes 3.x / Mantine 7 / Chakra v3 / Ark / shadcn. TOC actualizado. - eidos/README.md: tabla de referencia ampliada con §19. - CLAUDE.md: hand-off "2026-05-27 #6 (variants canon)". - CONTINUE.md: nota de la decisión arquitectónica. Variants component-specific permitidos (Banner inline/overlay/ persistent, Spinner bars/dots/ring, Button 'plain'): viven en cada `components/{c}/types.ts` y el lint los valida contra la CSS del componente. Tests: 101/101 pass en `src/uix/eidos`. `npm run check`: los mismos 6 errores pre-existentes (lib/_demo, soma/components/internal, web/ routes/active) — no relacionados. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	d3f9be4ab0	fix(toggle): add missing per-color track/hover/border forwarders ... The toggle palette cascade (palette-track / palette-hover / palette-border declarations in TSC) referenced --toggle-{X}-track, --toggle-{X}-hover, --toggle-{X}-border for X in {primary, neutral, secondary, affirm, risk, threat} — but the recipe only declared 5 forwarders per color (element / solid / solid-hover / text / contrast). The 3 missing slots × 6 colors = 18 unresolved tokens made --toggle-palette-{track,hover,border} resolve to empty for any non-neutral color (the host default DID declare them). Visible effect, only when a non-neutral color was active: - ghost-on-bg = var(--toggle-palette-track) → empty → transparent fallback - ghost-on-border = color-mix(... palette-border ...) → invalid → property dropped, border falls through to base layer - outline-on-bg = color-mix(... palette-element ...) was the one that worked because palette-element forwarders DID exist for all 6 colors Surfaced today while verifying the TSC v2.2 toggle-group color cascade fix in the browser. The toggle-group composition correctly overrode --toggle-palette-* on items, but the empty track/hover/border slots collapsed ghost-state styling downstream. Verified with browser probe (preview_inspect): - ghost + affirm: paletteTrack #0d1514, paletteHover #023b37, paletteBorder #145750 (was "") - ghost + threat: paletteTrack #191111, paletteHover #500f1c, paletteBorder #72232d (was "") - outline + primary: paletteTrack #18111b, paletteHover #3d224e, paletteBorder #54346b (was "") Tests: 100/100 pass in src/uix/eidos. Recipe contract guard accepts the new forwarders (kebab-case, valid var() values). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	4 months ago
dev	9c6375c96e	feat(eidos): TSC v2.2 (parts + composition) + universal theming coverage ... Token Scope Contract universal — no más excepciones arquitectónicas. Los 3 componentes que vivían fuera de TSC v2.1 (select, avatar, toggle-group) ahora están dentro del contrato via dos extensiones nuevas. TSC v2.2 extensiones (lib/config-types.ts + render-css.ts + config.ts): - `parts: readonly string[]` en RecipeTokenMultiDeclaration — emite selectores comma-separados (`[data-{c}-x], [data-{c}-y]`) para componentes con data-color cascadeado per-part. Consumer: select. - `composition: { foreignRecipe: { targetSelector, tokens } }` sibling key — overrides cross-recipe scoped a la cascade del host. Consumer: toggle-group modifica `--toggle-palette-*` en sus items. Migraciones: - select: 3 private `_accent-{track,border,text}` con parts: ['trigger', 'content'] + 7 cascades color:X each. Removed orphan `_accent-solid` (CSS no consumía). - avatar: 6 tokens via composite scopes `['variant:X', 'color:Y']` con matrix helper inline. Badge usa parts: ['badge']. Reemplaza 24 bloques CSS × 2 partes. - toggle-group: composition block con 8 palette tokens × 4 colors. Reemplaza 4 bloques CSS per-color. Bug toggle-group post-composition (encontrado y arreglado): Tras la composition migration, el cascade del toggle-group seguía roto porque los tokens derivados (`--toggle-solid-on-bg`, `--toggle-outline-fg`, etc.) viven en scope `[data-toggle]`. El `[data-toggle-group-item]` es sibling (no descendant), así que `var(--toggle-solid-on-bg)` resolvía undefined. Fix: inlined derivation expressions directamente en `[data-toggle-group-item]` y sus variant cascades (solid/outline/ghost), referenciando palette tokens en su propio scope local. Validador + emisor + contract: - `validateRecipeComposition` valida el shape `{ targetSelector, tokens }` y rechaza composition entries con scope='root'. - `stripCompositionKey` + `emitComposition` separan el pipeline. - `appendRecipeContractTokens` skip-list para `composition` (no aparece como fake `--{c}-composition` knob). - `tokenKeys`/`tokenEntries` helpers en recipe-css-contract.test.ts filtran composition en todos los iteradores. Documentación: - THEMING.md §18 reescrito como "Cobertura universal de TSC". §7 extendido con subsecciones "Multi-part scope" y "Cross-recipe composition" + ejemplos completos. TOC actualizado. - eidos/README.md tabla de referencia ampliada con TSC v2.2 + §18. - CLAUDE.md gana hand-off "2026-05-27 #5" (TSC v2.2 + cobertura universal). - CONTINUE.md reescrito al estado actual de la sesión. Working tree también incluye sprint Words en paralelo (multiple authors): slash menu, find/replace regex, code language picker, table audit, toolbar family menu, code highlight engine. Tests: 786/786 pass en src/uix/{eidos,morfo,soma,sema}. `npm run check`: 6 errores pre-existentes (lib/_demo, soma/components/internal, web/routes/active) no relacionados. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago
dev	28bf7f8b49	feat(pickers): wire runtime.trigger('close') in 5 picker providers ... After the 2026-05-27 picker family polymorphic refactor, the morfos declared the polymorphic `close` event but the providers were inert — they just toggled `opts.open = false` without firing the event. This commit cables them, mirroring the Dialog/Drawer/Popover dismissWith pattern (book §5.3). **Affected providers** (5): color-picker, date-picker, date-range-picker, time-picker, time-range-picker. **Pattern** (consistent across all 5): - Private `triggerClose(cause)` helper maps: 'committed' / 'range-committed' → commit.save + fulfill 'cancelled' → emerge.close 'dismissed' → emerge.dismiss 'dismissed-outside' → emerge.dismiss - Writes `data-last-action` on the morfo's target part (calendar / clock / content) imperatively before the trigger fires so eidos can tint the exit animation per cause. - Called from commit() / cancel() / handleSelect / handleDateSelect / handleRangeSelect — the close paths the provider already had. The morfo target varies per picker: - color-picker → content - date-picker → calendar - date-range-picker → calendar - time-picker → clock - time-range-picker → clock Internal state (`opts.open = false`) sequence preserved — the trigger fires BEFORE the open flip so the perceptual signal stamps `data-event-*` on a node that is still rendered. **dismissed / dismissed-outside causes** are declared but not yet wired to specific gestures — the picker shell delegates escape / outside-click handling to its wrapping Popover, whose own polymorphic close fires independently. The cause strings remain in the helper for future use (e.g. if a picker grows its own escape handling). **Result**: - Pickers now emit perceptual signal on commit/cancel via sema engine. - Eidos `[data-last-action='X']` selectors get exit-tint values to act on (mirror of Dialog/Drawer/Popover). - Picker tests: 18/18 still pass. `morfo:check`: 0 issues. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago
dev	312c8740ca	refactor(date-picker): rename kind 'date' to 'day' for trio consistency ... Audit caught `data-kind="date"` violation — the date-picker / date-range-picker morfos declare `data-kind ∈ [day, month, year]` but the public `kind` prop accepted `'date' | 'month' | 'year'` and defaulted to `'date'`. Rename prop value `'date' → 'day'` so the trio is consistent (all named after the grid cell granularity). **Rationale (book-pure)**: - The trio `day` / `month` / `year` is internally consistent — all three name the granularity of the grid cell the calendar view shows. - `'date'` was awkward in the trio: month and year pickers also pick "dates", just at coarser granularities. - The morfo enum was already canonical (`['day', 'month', 'year']`); this aligns the prop with the contract. **Surface touched** (all `'date' → 'day'` for the kind value): soma components: - date-picker / date-range-picker / date-field / date-range-field: provider type (DatePickerKind / DateRangePickerKind), default value, test fixtures, types.ts JSDoc. demos (web/routes/uix/components/): - date-picker / date-range-picker / date-range-field: kind state, selector chips, doc table. `'date'` literals that DON'T refer to kind (kept untouched): - ValueProfile = 'date' | 'datetime' | 'zoned' (different concept). - SegmentLock = 'none' | 'year' | 'monthYear' | 'date'. - CSS `input[type='date']` (HTML attribute). **Result**: - `npm run morfo:check`: 1 issue / 1 component → **0 issues**. All 107 routed morfos validate against their demo DOM ✓ - `npm run check`: 0 errors ✓ Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago
dev	59c3bc96c4	fix(color-picker): emit data-state explicitly on Provider props ... Audit caught 'data-state not emitted' on the ColorPicker Provider. The morfo declares `data-state ∈ [open, closed]` with `value: v.stateRef('open')` — the runtime's `syncAttrs` can sync this from `dom.apply`, but only if the provider receives a `ref` on its opts (which the ColorPickerProvider doesn't — `ColorPickerOpts` extends `ProviderOpts` + `StateProps` without the ref slot). Rather than re-shape the opts to add ref + bind it in the wrapper, just emit `data-state` explicitly in the provider's props derivation, same form as the other six declarative attrs that the provider already materialises (`data-format`, `data-invalid`, `data-disabled`, `data-readonly`, `data-required`, `aria-invalid`). Matches the dialog/drawer/popover doctrinal pattern for disclosure attrs. After this commit `morfo:check`: 1 issue remaining (date-picker `data-kind="date"` — separate domain issue, addressed next). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago
dev	9f1466dc57	fix(avatar): data-color severity optional to acknowledge custom-color escape hatch ... Audit caught 4 'missing data-color' violations on the Avatar Provider — the demo includes consumers passing non-canonical CSS color strings (`#7c3aed`, `oklch(70% 0.15 220)`, `rgb(245 158 11)`) which the wrapper correctly routes through `data-color-custom=""` + an inline `--avatar-color-custom` CSS variable instead of the enum attr. The morfo enum can't represent arbitrary CSS color strings, so the `data-color` attr is legitimately absent in those cases. The contract should reflect that: severity becomes `optional`. Symmetric with the `Avatar.Badge` part which already declared this attr as optional. Behaviour unchanged. Only the morfo's severity flag changes. After this fix `morfo:check`: 6 issues / 3 components → 2 issues / 2 components. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago
dev	3542f3a730	fix(button): remove redundant data-state enum (book §1 purity) ... Audit 2026-05-27 caught 11 `data-state="idle"` violations across 3 disclosure trigger components (dropdown-menu, table.row-detail-trigger, tree-grid.expand-trigger). Root cause was composition: each used `<Button>` as its chrome via `{...props}` spread, and Button's morfo declared a redundant `data-state ∈ [idle, loading]` enum that the runtime emitted unconditionally — clobbering the outer trigger's `data-state="open|closed"` via mergeProps order. Doctrinal fix: REMOVE `data-state` from Button's morfo entirely. Button's loading state is already fully expressed by the `data-loading` presence flag (declared further down in the same morfo and consumed by eidos as `[data-loading]`, not `[data-state='loading']`). The 'idle' value was a redundant synonym for "not loading" — and the book canon favours "presence = active state, absence = rest" for binary states (cf. `data-disabled`, `data-loading`). **Book-pure rationale**: - Disclosure (family `emerge`): `data-state ∈ [open, closed]` — dialog, drawer, popover, accordion, dropdown-menu, etc. - Lifecycle (family `sustain`): `data-state ∈ [idle, loading, ...]` — image, avatar, button (until now), card, field. - Mixing the two vocabularies in a single attr is the incoherence this fix removes. - Button is interactive surface, NOT lifecycle: its "rest" is natural absence of `data-loading`, not a distinct `idle` state. **Result**: - `npm run morfo:check`: 16 issues / 6 components → **6 issues / 3 components**. Removed all 11 `'idle'` violations. - Composition pattern restored: `<Button>` no longer clobbers `data-state` on the host trigger. - Eidos CSS unchanged (already uses `[data-loading]`). - Tests: 803/801 unchanged (2 remaining failures are unrelated, in the user's words editor WIP). **Not fixed in this commit** (3 remaining `morfo:check` failures): - `avatar.provider`: missing `data-color` emission (4×) - `color-picker.provider`: missing `data-state` emission - `date-picker.provider`: `data-kind="date"` not in declared `[day, month, year]` enum. Each requires a separate domain decision. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago
dev	681e056102	fix(uix-audit): close 4 pre-existing issues found in system audit ... System-wide audit of `src/uix/*` + `src/arts/adom` (excluding demos and web/routes). Findings + fixes documented in `src/uix/AUDIT_REPORT_2026-05-27.md`. **Fixed**: 1. `src/uix/eidos/lint.test.ts`: "every eidos CSS component has a morfo file" test extended to walk `morfo/internal/` as well as `morfo/components/`. The picker-shell relocation (audit-codex P1 #5) moved picker-shell to internal but the test wasn't updated. 2. `src/uix/contracts.test.ts`: two pre-existing failures resolved. (a) Picker WithChild regex updated to accept the `WithChild<Omit<PrimitiveDivAttributes, ...> & {...}>` shape applied in the audit-codex P0/P1 picker ProviderProps refactor. (b) `collectPublicSomaComponentDirs` now skips `picker-shell` as an internal primitive (symmetric to `collectKnownMorfoDataAttrs` which now walks `morfo/internal/` to include the picker-shell data-attrs). 3. `src/uix/soma/components/command/command-provider.svelte.ts`: restored missing `createAttrs(commandMorfo)` import. Was the only ERROR in `npm run check` baseline. 4. `src/uix/soma/components/dialog/types.ts`: JSDoc comment updated to reflect polymorphic close shape (cosmetic alignment with the 2026-05-27 refactor). **Result**: - `npm run check`: 1 ERROR → **0 ERRORS** ✓ - `npx vitest run src/uix`: 5 failed → **2 failed** (both 100% in user's words/* WIP, not framework code) - Polymorphic close + persistence + a11ySemantic rollout: **zero drift detected** across morfos / providers / sema cascades / tests. **NOT actioned** (documented in report): - 6 pre-existing `morfo:check` failures (avatar / color-picker / date-picker / dropdown-menu / table / tree-grid). All require domain decisions (drop 'idle' state vs add it to morfo enum, etc.). - 1 navigation-menu data-state='visible' eidos selector drift. - ~14 README/`.md` files with stale Dialog/Drawer/Popover descriptions. - Words editor doctrinal cleanup (user's WIP). - Tooltip polymorphic alignment (optional, no consumer benefit today). Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago
dev	34e0064f2c	feat(toast): explicit persistence: 'transient' on signal-announce event ... Documents the deliberate choice to keep the announce event's perceptual projection transient even though the canonical book §6.2 table would suggest `untilFix` for `signal + risk` or `untilAction` for `signal + threat`. Rationale: the Toast component owns its own lifecycle via `Toast.Provider`'s auto-dismiss timer + user dismiss. The `data-event-*` projection is a one-shot pulse when the toast enters. Persistence of the toast surface itself lives in `data-state` (open/closed) on the Item, not in the signal projection. Found via audit of all `verb: 'warn' | 'alert' | 'notify'` events across `src/uix/morfo/components/*`. Toast was the only one without an explicit persistence declaration — annotated now so the choice is intentional rather than implicit-default. (`words.signal-invalid-input` also lacks persistence but it belongs to the in-flight words editor work tracked separately by the user.) Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago
dev	9ca8b4e30b	docs(claude/libro): document persistence + a11ySemantic + polymorphic close rollout ... LIBRO_VARIACIONES_Y_EXTENSIONES.md: - New section D.9 (persistence + holds-by-intent — book §6.1). - New section D.10 (a11ySemantic per event — book §9.1). - New section D.11 (polymorphic events — book §5.3) with the full rollout state across overlays + picker family. CLAUDE.md hand-offs: - 2026-05-27: initial sprint (persistence + a11ySemantic + polymorphic types). - 2026-05-27 #2: caller messages + persistence extension + Dialog refactor. - 2026-05-27 #3: Drawer + Popover polymorphic refactor. - 2026-05-27 #4: Picker family polymorphic refactor + test fixtures decoupling. Each hand-off records: scope, files touched, doctrinal decisions, and pending follow-ups. The whole rollout was verified on 680/680 tests across sema + morfo + soma + adom scopes. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago
dev	8dce794ab4	refactor(pickers): collapse close-* into polymorphic close on 5 picker morfos ... Mirror of the dialog/drawer/popover refactor (book §5.3) applied to the picker family: - color-picker - date-picker - date-range-picker (incl. close-range-commit cause) - time-picker - time-range-picker Each had 4 close-* events with per-event prewrite; collapsed to a single polymorphic `close` event: semantic: { family: 'emerge', verb: 'close', target: v.partRef('calendar'|'clock'|'content'), sequence: 'pre', persistence: 'transient', allowedFamilies: ['emerge', 'commit', 'signal'] } **Behaviour preserved**: the picker providers don't currently fire any `runtime.trigger('close*')` — they just toggle `opts.open = false` and let the popover unmount. The previous `close-*` events were declared but inert (only consumed by schema validator + compiler tests). Refactor is doctrinal alignment, not behavioural change. **Sema cascade**: only color-picker has a sema pack and it did not reference any close-* event (only handle-pick / handle-drag / commit-set). Zero cascade updates needed. **Test fixtures**: compile.test.ts + runtime.svelte.test.ts already migrated to the synthetic `prewriteFixtureMorfo` in the core commit, so this refactor doesn't break them. After this commit the only morfo with the pre-polymorphic per-event prewrite shape is the synthetic test fixture itself — every production morfo is on the polymorphic close. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago
dev	9f5fbfdb71	refactor(dialog/drawer/popover): collapse close-* into polymorphic close (book §5.3) ... Three overlay components each had five `close-*` events (close-save / close-cancel / close-dismiss / close-dismiss-outside / close-after-fail) with per-event prewrite arrays setting `data-last-action`. Per the book §5.3 polymorphism canon, collapse to a single `close` event: semantic: { family: 'emerge', verb: 'close', target: ..., sequence: 'pre', persistence: 'transient', allowedFamilies: ['emerge', 'commit', 'signal'] } NO prewrite — the provider sets data-last-action imperatively before trigger because polymorphic events can't bind a single prewrite per call. **Provider pattern** (Dialog / Drawer / Popover): - New `DISMISS_CAUSES` map: { save, cancel, dismiss, dismiss-outside, fail } → { lastAction, semantic: { family, verb, intent? } }. - `dismissWith(action, opts?)` does `dom.apply({ 'data-last-action': cause.lastAction })` then `runtime.trigger('close', { fallbackTarget, semantic: cause.semantic, message })`. - `triggerClose` is now private; previously-public 2-arg signature collapsed into the cause-driven internal helper. - All internal callsites (escape, outside-click, hover-close, close button) migrated to `dismissWith`. - `opts.message` forwarded to runtime — when `a11ySemantic.requiresLiveRegion` is set (today only for the 'fail' case), the live region announces it. **Schema validator relaxed** (`src/uix/morfo/schema.ts`): the "every declared value of `data-last-action.values[]` must be prewritten by some event" check was DROPPED — provider-imperative writes can't satisfy it. The reverse direction (`prewritten ⊆ declared`) stays strict. **Cascade rewrite** (`src/uix/sema/components/{dialog,drawer,popover}.ts`): - `eventNamePrefix: 'close-'` → `eventName: 'close', eventFamily: 'emerge'` - `eventName: 'close-dismiss-outside'` → `eventName: 'close', state: { attr: 'data-last-action', value: 'dismissed-outside' }` **Eidos CSS**: not touched — already reads `data-last-action` directly. **Public API preserved**: `dismissWith` signature + observable behaviour identical. External consumers see no difference. Tests updated for the new event name. 680/680 pass across sema + morfo + soma + adom scopes. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago
dev	c34fd1fd31	feat(uix-components): apply persistence + a11ySemantic + live-region messages to 6 morfos ... Per the canonical book §6.2 + §9.2 mapping, six morfos with `signal.warn`, `signal.alert` or `signal.notify` events get explicit `persistence` + `a11ySemantic` annotations, and their providers wire up `clearTarget` + caller-message integration with the shared live region. **Morfo annotations**: | Morfo / event | persistence | a11ySemantic | |--------------------------------------------|----------------|---------------------------------------------------------| | announce.signal-alert | untilAction | requiresPersistentTrace + requiresLiveRegion | | form.signal-warn-invalid | untilFix | requiresPersistentTrace + requiresFocusMove + reducedMotionFallback='text' | | file-upload.signal-warn-reject | untilFix | requiresPersistentTrace + reducedMotionFallback='text' | | password-field.signal-notify-caps-state | stateBound | requiresLiveRegion + reducedMotionFallback='text' | | tags-input.signal-warn-reject | untilFix | requiresPersistentTrace + reducedMotionFallback='text' | | textarea.signal-warn-count-overflow | untilFix | requiresPersistentTrace + reducedMotionFallback='text' | **Provider wiring**: - form-provider: clearTarget on submit, message="{N} form errors..." via FORM_LANGS.ERROR_SUMMARY_*. - file-upload-provider: clearTarget before each accept/reject cycle, message via new FILE_UPLOAD_LANGS.REJECT_SUMMARY_{SINGLE,MULTI}. - password-field-provider: clearTarget on caps-off transition, message via existing PASSWORD_FIELD_LANGS.CAPS_WARNING. - tags-input-provider: helper `emitWarnReject` (clear + emit + message via new TAGS_INPUT_LANGS.REJECT_WARNING), called from addTagValue, addFromInput, handlePaste. Successful add also clears (the add is the fix). - textarea-provider: emits only on transition INTO overflow, clears in transition OUT (untilFix projection persists). Message via new TEXTAREA_LANGS.OVERFLOW_WARNING with {{max}}. Three new lang catalog entries (reject/overflow warnings) merged into the existing components.{name}.* namespaces. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago
dev	70f6f3bf0d	feat(uix): persistence + a11ySemantic + polymorphic core (book §5.3, §6, §9) ... Three book canon items codified in code: **Persistence (libro §6.1)** — separa `hold` (perceptual min) de `persistence` (lifecycle real): - New `SignalPersistence = 'transient' | 'untilAction' | 'untilFix' | 'stateBound'` - New `SEMA_HOLDS_BY_INTENT` canonical lookup table (`src/uix/sema/holds.ts`) - `EngineSemantic.emit()` returns the resolved signal id; keeps projection alive past hold for non-transient. Exposes `clear(id)`, `clearTarget(target)`, `hasActive(id)`. - `SomaRuntime.trigger()` returns `TriggerResult { id?, persistence? }`. Exposes `clearSignal(id)`, `clearTarget(target)`, `partRef(part)`. **a11ySemantic (libro §9.1)**: - New `MorfoA11ySemantic` (requiresPersistentTrace, requiresLiveRegion, requiresFocusMove, keyboardEquivalent, reducedMotionFallback) on MorfoEvent. - `ActiveDom.prefersReducedMotion`: reactive tracker via media query (`src/arts/adom/reduced-motion.svelte.ts`). - `ActiveUix.announce(msg, priority?, timeout?)`: lazy-created live region via `dom.writeNode`. - `Soma.runtime()` auto-wires `sources.announce`. - `SomaRuntime.trigger()` honors a11ySemantic after emit (live region, focus, reduced-motion fallback including `channels: []` for 'state'). **Polymorphic events (libro §5.3)** — ADITIVO sobre shape concreto: - `allowedFamilies?: readonly SemaFamily[]` opcional en MorfoEventSemantic. - Provider override via `runtime.trigger(name, { semantic })`. - Default family of the morfo is implicitly allowed. - New `SomaRuntimePolymorphicError`. - New `isPolymorphicSemantic` type guard. Synthetic `prewriteFixtureMorfo` (src/uix/morfo/test-fixtures.ts) decouples compile + runtime tests from the production morfo catalogue. 671 tests pass in sema + morfo + soma + adom scopes. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>	5 months ago