Formato puro: `npm run format` (dos pases; siete ficheros no convergían en el
primero) sobre todo lo que .prettierignore no excluye (web/ congelado,
.claude/, artefactos generados). Este commit va en .git-blame-ignore-revs.
Neutralidad medida fichero a fichero, compilando y minificando con esbuild (y
con el compilador de Svelte, cliente y servidor, para .svelte) la versión de
HEAD y la formateada: 1 031 ficheros de código idénticos; 9 CSS que solo
difieren en espacios pegados a un paréntesis (`var( --x )` frente a
`var(--x)`), que CSS no tokeniza como significativos; 510 Markdown sin
compilación posible (docs:check los valida). Los dos generados y el guard que
dependía del formato se resolvieron en el commit anterior.
Verificación sobre este árbol: npm run lint exit 0 · gate entero verde salvo
focus-census, arreglado en el commit anterior y re-ejecutado (9/9): check:gate
OK (src/ y scripts/ a cero) · docs:check 0/0 · arts/blocks/packs/rtl/
translations/agent · eidos:lint 0 inválidos · apps:check (boot 27/27, 0
errores) · suite 465/466 → 466/466 con el guard corregido · build de la raíz
exit 0.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
La entrada eidos se parte igual que la de soma: `activeEidosDir(dir)` devuelve
la AFIRMACION (prop -> ancestro) y publica en el MISMO DirectionContext — un
chart eidos-only dentro de un subarbol soma afirmado (o al reves) resuelve el
mismo hecho. La cola matematica va a `resolveEidosDir(dir, eidos.prefs)`, con
la vista de prefs cacheada por instancia (WeakMap): construirla dentro de una
funcion pura llamada por $derived era una alocacion por pasada.
createChartRtl consume las dos mitades por su lado: `attr` estampa la
afirmacion cruda (ausente hereda del <html> proyectado), `current`/`anchor`
resuelven la cola.
Corpus:
- direction-contract.md §2 pasa del campo obligatorio de 4.1 al mecanismo del
morfo (declaracion -> tipo condicional -> estampado; secundarios del mismo
morfo; el censo como guard), §7 documenta el contexto compartido y la
tabla de entradas partidas, y el checklist §8.3 pide morfo + cable.
- docs/decisions.md registra la ratificacion D1-D4 del 2026-08-05 con las
cuatro decisiones y su porque.
- CONTINUE-direction-runtime.md §11 cierra el handoff: tabla de fases con
commits, verificacion final medida, la cola pospuesta por D4 y las trampas
nuevas (pathspec SIEMPRE en rama compartida; el detector de huerfanos y la
ruta del import; la vista en $derived).
check 77 = linea base · morfo+direction suites 131/131 · docs:check 0/566.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Decision del usuario: escribir el contrato antes de tocar codigo.
El documento existe porque `settings-button` lleva declarada desde el principio
prometiendo «velocidad / calidad / pista», y al ir a implementarla resulta que
dos de los tres ejes ya no le corresponden y el tercero no existe.
Auditado sobre el codigo: grep de `quality|bitrate|resolution|level|
representation` sobre el puerto da CERO. Ni `MediaSnapshot` (0 de 14 campos), ni
`MediaProvider` (0 de 9 miembros), ni los espejos de `$arts/sound`.
Y no es un descuido: la calidad NO es intrinseca al `<video>`. Un fichero no
tiene niveles; los niveles aparecen con un manifiesto HLS/DASH y un motor que
elige representacion. Este framework no vendoriza motores, asi que la lista solo
puede venir del consumidor. De ahi que el contrato, si se abre, sera una lista
VACIA para el 90 % de los usos — lo que decide la forma: el control se AUSENTA,
no sale vacio. Misma doctrina que el waveform sin picos.
Queda escrito el contrato minimo (3 campos de snapshot con `qualityId` separado
de `qualityAuto`, `setQuality`, los dos espejos, el evento de morfo que rompe el
censo de 11, los textos) y la trampa que ya nos mordio con la velocidad: el
commit cabalga el RESULTADO, no el clic — pedir 1080p no es haberlo conseguido.
Registrado en el indice `docs/decisions.md`. docs:check 0.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
El eje estaba cerrado en CODIGO y no en DOCUMENTACION. El contrato vivia solo en
`CONTINUE-direction.md`, un handoff que `docs/README.md` declara «never a source
of truth». Un capitulo E2 lo fija ahora: `docs/canon/direction-contract.md`.
Va a canon y no a arquitectura por la misma razon que `recipe-contract.md`: es
normativo y tiene guard ejecutable. Cubre la cadena y donde corre cada eslabon;
por que `undefined` no es `'ltr'`; las DOS atributos —`dir` crudo (nativo, lo que
`:dir()` mira) y `data-dir` resuelto (opt-in, para recetas que necesitan un hook
incondicional)—; cuando el estampado es OBLIGATORIO; la doctrina de selector; las
trampas que el guard no ve; que espeja y que no; y la mitad global de prefs.
EL CORPUS SE CONTRADECIA en cuatro sitios, y dos de ellos ENSENABAN mal:
- `component-guide.md:372` daba la cadena como `prefs → 'ltr'`, DOS eslabones,
mandando al wrapper a leer prefs directamente. Eso excluye la prop.
- `soma-architecture.md` §3.4 presentaba el resolutor a pelo (`soma.prefs.getDir()`
dentro del provider) como LA forma de obtener la direccion — justo lo que el eje
retiro del catalogo.
- `html.ts` ensenaba en su JSDoc `dir?: 'ltr' | 'rtl'`, el union a mano que la
regla E-2.5 ahora prohibe. El mecanismo que ensena (extender para estrechar una
clave HTML) es correcto y sobrevive; solo cambia el ejemplo.
- `active-architecture.md:416` no listaba `lang` en la proyeccion, contra
`contracts.ts` y contra su propio test de frontera. Igual `prefs/README.md`.
Y no lo mencionaba en absoluto: el glosario (`activeDir`, `resolvedDir`,
`data-dir`, RTL-1, el escape `rtl-physical:`), `eidos.md` (RTL-1 es el TERCER
guard de deriva y faltaba junto al builder tipado y eidos-lint), `soma.md`,
`building-a-component.md` (§Known traps es exactamente donde va «la prop mueve la
matematica y deja la pintura atras»), y la tabla de atributos de la arquitectura,
que afirma «todo lo que viaja entre capas viaja por atributos» y omitia `dir`.
El checklist gana cuatro reglas y el guard que faltaba (`rtl:check` no estaba en
la matriz de aceptacion pese a que la tabla canon lo nombra), mas un recorte en
E-3.5: «todas las props visuales mapean a `data-{prop}`» leia como mandato de
estampar `data-dir`, que `:dir()` no puede ver.
DIVERGENCIA DECLARADA (§7): la familia `chart` no tiene prop ni provider y
resuelve leyendo `getComputedStyle(node).direction`. Es el unico sitio del
catalogo que hace lo que §1 prohibe. Se registra en vez de esconderse.
Tres verificadores adversariales sobre el barrido; sus hallazgos, corregidos:
RTL-1 estaba anunciado como guard de TODO el capitulo (solo cubre §4) · «nunca
lee al padre» era absoluto y borraba la composicion sancionada en el punto de
llamada · el estampado se afirmaba incondicional en un sitio y condicional en
otro · las cuatro filas nuevas usaban una aplicabilidad que ni la leyenda ni
`component-audit.ts` conocen.
`docs:check` 0 errores sobre 564 docs. `check` 77 = la linea base.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Documentación del bloque chat-* recién shippeado, siguiendo la convención
del corpus (por-componente = READMEs dobles; por-lib = README; doctrina de
bloque = design record en decisions).
- docs/decisions/design-chat-block.md: la doctrina cross-cutting que abarca
los 4 componentes — la tesis comparativa (5 fallos del sector = nuestras
oportunidades), la topología de composición (triple registro Feed +
VirtualList anclado, Textarea, FileUpload, Icon lucide), el anclaje
(sticky monótono, pin contra DOM real), la doctrina visual del rediseño
de referencia y los 2 fixes de framework que destapó. Indexado en
decisions.md (sección "Component families — design records").
- src/libs/chat/README.md: los helpers puros (groupIntoRuns ventana 8min,
aggregateTypers, límites de día) — convención de README por lib.
docs:check 0 errores (503 docs). Los READMEs por componente y next-features
§7 ya iban en el commit del bloque (4c4878f13).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Two streams, split by what the animation touches:
STREAM A — decorative backgrounds → the pack tier
- arts/scene: a consolidated scene runtime ($scene) that owns, once, the
citizenship every ad-hoc background reinvented or skipped (frame loop,
off-view pause, DPR cap, mandatory reduced-motion, WebGL context
loss/restore, scene budget, teardown). SceneDom port (adom satisfies it),
webgl/webgl2/canvas2d drivers + a custom-pipeline extension (vertexShader +
draw + glContext.depth/dprCap) for real geometry (beam, particles, dither,
grid, eter, pixel-blast, hyperspeed). 32 effects as shared resources.
- src/packs/ambient: the first pack — <Ambient effect="…"> mounts a registered
effect; the P contract (P-1..P-6) guarded by scripts/packs-check.ts; colors
are token-aware (P-4). One-way dependency, removable-by-construction.
- resolveToken extended to semantic color slots (--color-{role}-{slot}) so
consumers resolve theme tokens to concrete colors (the P-4 half).
STREAM B — animations over real text → canon
- Six components (count-up + text-{gradient,circular,blur,focus,scramble}):
each a morfo + eidos recipe (where there's styling) + demo. CountUp is a
service component (counts through uix.format.numbers). The five Text* are
passive decoratives. Upgrades over the seeds: SR hardening (real text
visually-hidden + aria-hidden decoration), a11y fix (no fake role=button),
measurement discipline (cached rects via dom.measure, no reflow storm),
reduced-motion, ecosystem citizenship (eidos.dom/timers, no raw platform).
- MorfoElement gains 'p'.
DOCS
- docs/architecture/packs.md (pack tier, admission rule, P contract, Aura
promotion path); docs/decisions/design-text-effects.md (the family design
record) + indexed in decisions.md / README.md; glossary entries
(scene/Ambient/Aura/text effects); scene README custom-pipeline + authoring
bridge; motion-guide content-effects note; strata tables acknowledge packs.
Gates: component:audit 141/0/0 · docs:check 0/0 · scene tests 9/9 · packs:check
0/36 · check 0 own errors. Verified in browser (32 effects mount+compile;
6 text components SSR+hydrate, CountUp re-formats by locale live, TextGradient
resolves token stops to OKLCH via var()).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Move the three superseded arts design-phase records into the E3 corpus
(already indexed in docs/decisions.md), verbatim-as-historical:
- docs/decisions/design-connection.md (from src/arts/connection/DESIGN_CONN.md)
- docs/decisions/design-session.md (from src/arts/session/DESIGN.md)
- docs/decisions/design-timer.md (from src/arts/timer/DESIGN_TIMR.md, kept Spanish)
Each gets a `status: historical` frontmatter block; the current reference for
each subsystem is its README.md. A stub at every old src/arts/* path keeps the
~30 source provenance citations (`DESIGN_TIMR §12.8`, `DESIGN.md §7.Q1`, …)
resolving. Bodies kept verbatim (no reformat). docs/decisions.md index + naming
note updated to the new paths. docs:check 0 errors.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
User-caught drift, and it was live in the UI, not just prose: the demo
harness's hand-copied PALETTE_GROUPS enumerated 31 scales — fuchsia and
steel (the two non-Radix scales) never appeared in any demo's
PalettePicker even though PALETTE_SCALES (the const, 33) and the recipe
cascade (paletteScaleDecls maps the const) resolve them fine. Fixed the
two missing entries, typed the groups as PaletteScale[], and added a
dev-time completeness assertion so growing the const without grouping
throws at module load.
Prose swept (the numbers were inherited verbatim from the pre-book
sources, which contradicted their own '33-scale library' headers):
decisions.md color-model row, demo-authoring x4 (incl. the '31-scale
palette' lede), rfc-color-engine x4 (ground-truth + templates
open-question). Where the count was incidental the text now points at
PALETTE_SCALES instead of restating a number.
Guard: docs-check I1 gains the palette count (EN + ES patterns,
~-approximation and type/space-scale exclusions; s-mark added to the
section-number skip). The check line now prints palette=33. This was
the exact class I1 already covered for families/intents/archetypes and
had not been extended to the palette.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The global sweep caught the last corpus links still pointing at
pre-book paths: morfo chapter's s2.bis cross-ref (src/uix/README ->
overview.md), authoring's E1 stratum row (active_architecture ->
docs/architecture/), the related: frontmatter of decisions.md and
canon/recipe-contract.md (THEMING/THEMING_GUIDE -> theming/reference +
guide), and component-audit's typographic-vertebration pointer (eidos
README -> architecture/eidos.md with the translated heading). README
map completed as the book TOC: every chapter now has a row —
theming/motion.md (the model) and theming/changelog.md (the chronicle)
were the two missing. Remaining old-path references live only in
chronicle docs (theming/changelog, process/) and resolve via stubs;
the 11-warn baseline is all foreign targets (MOTION_SERVICE_RFC demo
routes, deleted THEMING_AUDIT/PENDIENTES).
F7.6 complete. The book tree is fully populated: architecture/ (8) +
canon/ (2) + theming/ (7) + rfcs/ (7) + guides/ (4) + decisions/ (2).
Next and last: F7.7 CLAUDE.md thinning — diff to the user first.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
COLOR_ENGINE_RFC (619 L, the largest RFC) -> rfcs/rfc-color-engine.md,
Spanish to English, same section numbering (s0 TL;DR, s6 generator +
s6.1 isomorphism + s6.2 deriveScheme/buildScheme, s7 wide-gamut
strategies, s8 APCA, s10 frozen invariants, s11 phases with 4-bis
implemented, s13 bloat-doc absorption, s14 comparison). Its
THEMING_AUDIT citation notes the audit is retired; the s25/s26 canon
citations point into theming/reference.md. decisions.md row repointed +
Status aligned (implemented through phase 4-bis); the index's naming
note rewritten — the deferred eidos-RFC rename is now DONE via stubs,
while the arts DESIGN_* docs keep their legacy names in place (their
citations were not swept).
F7.4 complete: 7 RFCs live at docs/rfcs/rfc-{scaling, structure, depth,
shape, typography, color-model, color-engine}.md; MOTION_SERVICE_RFC
stays in src (foreign/concurrent). docs:check 0 errors.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
TYPOGRAPHY_ENGINE_RFC (191 L) -> rfcs/rfc-typography.md and COLOR_MODEL_RFC
(155 L) -> rfcs/rfc-color-model.md, Spanish to English. color-model is a
RESOLVED RFC kept as the historical record of the rejected single-anchor
proposal — the resolution header travels with it, canonical model stays
theming reference s25. typography's stub notes the measurement arm
(lib/canvas-text, SText) stays in src. decisions.md rows repointed;
typography's Status aligned with the RFC's own closed-phases state.
Theming reference s21/s25 links swept. docs:check 0 errors.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
DEPTH_ENGINE_RFC (149 L) -> rfcs/rfc-depth.md and SHAPE_ENGINE_RFC (162 L)
-> rfcs/rfc-shape.md, Spanish to English. Both stubs summarize the shipped
state (phases 1-5) and keep the by-path citations resolving — including
the '// (DEPTH_ENGINE_RFC s5)' comments across src/uix/eidos/lib/*.ts that
decisions.md documents as load-bearing. decisions.md index rows repointed
and their Status column aligned with the status the RFCs themselves
declare (both said implemented; the rows still said Propuesta). Theming
reference s29/s30 canonical-guide links swept. docs:check 0 errors.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
SCALING_RFC (92 L) -> rfcs/rfc-scaling.md and STRUCTURE_ENGINE_RFC (113 L)
-> rfcs/rfc-structure.md, Spanish to English, same section numbering. The
legacy-name -> rfc-* rename that decisions.md declared deferred 'until
those citations are swept in the same pass' happens here: stubs at the old
paths keep every by-path citation resolving. Corpus links swept (theming
reference s23/s31, decisions.md index rows, channels.md capstone list —
the channels.md sibling links to rfc-color-engine/typography/depth/shape
land later in this batch). Chronicle docs (theming/changelog,
inherit_audit) keep old paths by doctrine — the stubs resolve them.
docs:check 0 errors.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
src/uix/active_architecture.md (865 L, Spanish) translated to English as
docs/architecture/active-architecture.md: the s0 module contracts +
canonical naming, the four layers in depth, the transcription chain, the
causal chain of one interaction, the DOM-attribute table, the 12 hard
rules (incl. the sec-dom read-timing rule), authorship vs transcription,
what it is NOT, risks. The dated header changelog block was dropped per
the timeless rule (s10 already routes snapshots to docs/process/); the
stale themes/fonts.css mention in the eidos structure listing corrected
to the generated foundation (phase-6 fact). Stub at the old path carries
the full sN section map (the most sN-cited doc in the corpus). All docs/
corpus links repointed to the chapter. docs:check 0 errors.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
decisions.md (the E3 design-rationale entry point) gains a cross-cutting entry
for the read-timing & token-resolution work — including the decision to reject a
static grep-guard in favour of the runtime `uix.perf` detector — pointing to where
the full argument lives (active_architecture §7 + arts/adom + arts/perf READMEs).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
A single E3 entry point cataloguing the design rationale: the 7 eidos RFCs
(color model/engine, typography/depth/shape/structure engines, scaling), the
3 arts subsystem design docs (connection/timer/session), and the cross-cutting
decision logs (LIBRO_VARIACIONES, GESTURES). Each entry gives status + the one
decision it records, linking the document for the full argument.
This delivers the "naming único" goal at the index level. The physical file
rename (*_RFC -> rfc-*, DESIGN_* -> design-*) is deferred: those names are cited
as provenance anchors in ~30 source files (eidos/lib/*.ts, arts/timer/*,
arts/color/*, tests), so a rename only pays off if every citation is swept in
the same pass. The index gives consistent naming without that churn.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
dev