Opus 5 turned the iterate-with-screenshots-until-it-meets-the-bar
instruction into 42 screenshot trips and 150 tool calls per build,
about forty dollars of cache churn a page, before ever reaching the
reviewer. Verification now batches: one desktop-and-mobile round after
the full build, fixes applied together, one confirming round, ceiling
two. Craft-floor's checks share those renders instead of earning
separate trips; per-tweak iteration is live mode's channel.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
From the paired Opus and Codex manual-run analyses. DESIGN.md moves to
the end of the flow and into a shipped documenter subagent that derives
the system from the built artifact: a rulebook written before the build
gets defended against reality, and a half-stable DESIGN.md hands the
design-system detector an unstable target that buries the build in
noise and invites laundering. The finish reviewer gains the handoff
that failed three times live: the parent captures desktop and mobile
screenshots and passes paths, the reviewer never attempts to render
and names missing inputs in one line, the parent verifies the
five-section return and respawns once on empty. Fidelity against the
approved comp joins its checks; the card keeps commitment only. The
comp ingredient inventory becomes a written gate with raster-by-default
materials and no gradient-as-texture, comps persist under
.impeccable/mocks, the degraded seed names the sandboxed-exec cause,
and the finish line is explicit: a clean detector pass is not finished.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The release-gate audit traced four ways the roll's output was defeated
downstream of a perfectly healthy seed. Gemini's harness keeps only the
tail of tool output, so the header-only ASSIGNED INDEX never reached
the model in 18 of 18 samples; the seed now restates the assignment
and key at the end of its output. Astro strips frontmatter comments,
so half the anthropic contracts vanished from built artifacts; the
contract now must survive the production build as an HTML comment in
emitted markup. A brief that paints its own picture (the album named
Soft Cathedrals) converged every arm regardless of assigned index; its
literal reading now joins the rut with at most one candidate. And Opus
under 4.0.1 skipped the seed 42% of the time while hand-authoring
plausible contracts; the finish reviewer now verifies FORM carries a
corroborable seed key before any craft point.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul's call: the build-exhaustion failure only exists inside eval
workers with hard turn budgets no real harness exposes, and the clause
doubled as a hedge door for skipping the comp round. The eval-side fix
belongs in the worker's max-turns, not in skill prose.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The release-gate campaign confirmed two skill bugs with transcripts.
Sandboxed harnesses reject absolute paths, so following the CHOSEN
CARD directive with the absolute card-base path failed view_image; the
directive and the quality-bar clause now say download into the
workspace and open the relative path. And under the openai worker's
turn cap, models spent the budget on init, cards, and comp generation
and never built the page (a third of small-n supplement slices); the
visualize mandate gains its one exception: at a hard cap the shipped
page outranks optional imagery.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul's call: the reviewer-local authorization patch covered one
command while the harness gate silently disables every shipped
subagent, critique panels and the manual-edit applier included. The
argument now lives beside the autonomy counter in context.mjs, emitted
as tool-result content every run: invoking the skill is the user
request such gates ask for; spawn where a reference directs; the
in-thread substitute is for absent capability only and gets disclosed
in one line. new-work keeps the reviewer mechanics and drops the
now-central argument.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A live session on a harness whose guidance gates subagent use on user
request resolved the conflict silently against the skill: it never
spawned the reviewer, stretched the no-subagents fallback to cover
permission hesitancy, and self-reviewed with all the context that made
its choices feel correct. Three tightenings: invoking the skill IS the
user request that authorizes its shipped subagents; the fallback is
for harnesses lacking the capability, not for hesitancy; a substituted
review gets disclosed in one line at finish, never silently.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Three browser-side fixes for the same 3.5-to-4.0.1 regression.
- Source-preview targets no longer source-inject per variant_progress
checkpoint. Immediate injection raced framework (React/Vue) ownership and
triggered removeChild errors, which surfaced as static previews. HMR now
owns reconciliation while variants stream in; source injection runs only on
the final done (its 750ms settle + retry ladder stays for non-HMR harnesses
like Cursor). Progress counts still advance from the variant observer, and
the svelte-component progressive path is unchanged.
- The agent-phase progress bar advances monotonically. A behind/resumed
checkpoint re-broadcasts an earlier phase (the server regresses the snapshot
phase to generating), which moved the visible bar backward; a phase rank
table now blocks a known-lower phase from overwriting a known-higher one.
- The server-lost toast now frames the drop as resumable (session saved,
reopen or restart live-poll.mjs) instead of "Session ended", which had led
agents to rationalize bailing to direct edits.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The polling-rework preflight wrote the variant scaffold into source during the
poll lease, before the agent acted. On source-preview targets (React/Vue/Vite,
everything but the svelte-component path) that write full-reloaded the
framework; a browser caught mid-reload missed the agent's variant write and the
SSE done, and sat stranded at 0/N.
Restore the 3.5 single-atomic-edit semantics: the preflight still resolves the
element location and computes the scaffold, but --defer-source-write leaves
source untouched and hands the agent the wrapper text plus the picked source
range. The agent splices variants into the wrapper and replaces the range in
one write, so the framework reloads exactly once. The svelte-component path is
untouched (it never writes route source). The missed-completion recovery stays
as defense in depth.
Also cache the resolved source file per target signature (locator + route):
the ~7.6s tree search re-ran on every generate for the same element; a hit now
points the helper straight at the file via --file, invalidated when the target
changes or a resolution fails.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Live mode had no TanStack coverage: a TanStack Start user hit disconnects
and static previews because there is no static index.html to inject and no
adapter for the SSR root document.
- New tanstack-adapter.mjs, modeled on the SvelteKit/Nuxt adapters: detects
a TanStack Start project (@tanstack/react-start + src/routes/__root.tsx)
and patches the __root document to mount a generated dev-only React
component (src/impeccable/ImpeccableLiveRoot) that appends the live bundle
on the client after hydration, carrying the ?token= param via
buildLiveScriptSrc. Patch/unpatch round-trips byte-for-byte and is
idempotent; refuses to clobber an unmanaged file at the component path.
- Wire detection into live-inject.mjs (insert + remove + gitignore),
ordered so SvelteKit/Nuxt win and a plain TanStack Router SPA falls
through to the baseline Vite index.html path.
- tanstack-router-vite fixture (baseline, no adapter) and tanstack-start
fixture (SSR adapter), both with runtime blocks. Both pass the full
live-e2e cycle (handshake, steer, pick, Go, cycle, accept, carbonize,
reloadProbe).
- Unit tests for detection + patch round-trip + apply/remove; tanstack-start
branches in framework-fixtures.test.mjs; live.md framework table + adapter note.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Two prose fixes from the 3.5-to-4.0.1 forensic diff of a real user
regression (15-minute tweaks, repeated disconnects, agent abandoning
the picker). The craft-fold made every generate cycle pay the verify-
the-built-result loop the overlay already provides to the human; live
cycles now verify by construction and run the full check once at
accept. And nothing framed a dropped SSE or closed tab as resumable,
while the client toasts "Session ended", so agents rationalized
bailing to direct edits; the journal is canonical and reopening
continues the session.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Two defense-in-depth layers close the P1 in issue #304, where any browser
tab on the machine could fetch /live.js, extract the embedded token, and
drive every token-gated route.
1. Loopback-restricted CORS. The shared handler replaced its wildcard
`Access-Control-Allow-Origin: *` with reflection gated on a strict
isLoopbackOrigin() that URL-parses the Origin (so localhost.evil.com and
127.0.0.1.evil.com fail) and accepts only http/https on localhost,
127.0.0.1, or [::1]. Reflection always pairs with `Vary: Origin` so a
cache never hands one origin's authorized response to another. Remote
origins get no ACAO header; origin-less callers (script tags, curl, the
agent's own fetches) are unaffected.
2. Token-gated /live.js. The handler now 401s unless `?token=` matches
state.token, so the bundle (which embeds the token) is no longer served
to unauthenticated local pages. The injected <script src> carries the
token: live.mjs passes --token to live-inject.mjs, which threads it
through every injection path (HTML/JSX tag, Nuxt plugin, SvelteKit root
component) via a shared buildLiveScriptSrc(). The token stays optional in
live-inject so static fixture tests keep their bare src.
Tests: new live-server integration cases for the 401 gate, remote-origin
denial, loopback reflection + Vary, and token-guarded routes under a
loopback Origin; e2e session harness now injects with the token.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The /source route confined paths with `absPath.startsWith(process.cwd())`,
a string-prefix check with no separator. An absolute request path to a
sibling directory whose name extends the project dir name (projeto ->
projeto-backup) shared the prefix and was served. Switch to the relative-path
check already used by sessionFileMetadataFromPollReply: reject when the
relative path is empty (the root dir itself, never a file this route serves),
starts with `..`, or is absolute.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The deep staleness pass extracted a hook-script path with a greedy `\S*`
prefix that swallowed the `${CLAUDE_PROJECT_DIR}/` placeholder, then
existsSync'd the literal string. That string never exists, so every project
installed by `impeccable hooks on` got a `hook-script-missing` finding with
text claiming UI edits were going unscanned — the opposite of the truth.
Split extraction from resolution. hookScriptTokenFrom now pulls the path
token (quoted-first, so it handles the #399 guarded `[ ! -f "PATH" ] || node
"PATH"` form and absolute user-level installs) without absorbing shell
syntax. resolveHookScriptPath then applies a per-placeholder policy:
- ${CLAUDE_PROJECT_DIR} expands to the scanned root (the runtime mapping).
- ${CLAUDE_PLUGIN_ROOT} / ${PLUGIN_ROOT} / ${GROK_PLUGIN_ROOT}, $(...) command
substitution (GitHub's $(git rev-parse)), and any other $VAR are SKIPPED:
the doctor cannot know those locations and must never assert a negative it
cannot verify.
The check stays real: a placeholder that expands to a genuinely absent path
still flags. Adds TDD coverage for every command form.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The Stop deep pass (runStopHook) never read the stop_hook_active field
from the Claude Code Stop-hook event. When a prior fire kept the turn
alive via hookSpecificOutput.additionalContext and the agent legitimately
declined to act, the hook re-scanned and re-blocked every re-invocation
until Claude Code's consecutive-block cap force-ended the turn (issue #400).
Read stop_hook_active early in runStopHook, right after the event is
parsed and before any scan, and exit 0 with no output when it is true. The
prior fire already surfaced the findings; acting on them is the agent's
call. Only Claude Code sends this field, so the strict === true is a no-op
for other harnesses. runHook (PostToolUse) and hook-before-edit.mjs
(PreToolUse) never receive the field, so they are unchanged.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The local-image map fell through to image/jpeg for anything that was
not webp or png, so an svg hero (the fake comp generator's native
format) silently failed to render on the decision page.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A cheap, LLM-free E2E tier for the interactive parts of new-work, mirroring
the two-layer live-e2e pattern (deterministic now, opt-in LLM tier later).
- generate-image.mjs: IMPECCABLE_IMAGE_GEN_FAKE=1 writes a deterministic
offline image (SVG with wrapped prompt + SYNTHETIC COMP label, or a valid
palette-stripe PNG carrying the prompt/marker in a tEXt chunk). Same CLI
contract, no key, no network, $0.00 cost line.
- tests/new-work-e2e/user-bot.mjs: scripted user bot (module + CLI) that
resolves the serve-question daemon from the workspace and drives the real
page via Playwright (pick, re-roll + steer, canon, tab close).
- tests/new-work-e2e.test.mjs: node --test coverage of the serve-question
cycles (pick + CHOSEN CARD, re-roll + --update re-deal, canon + CANON
CHOSEN, tab-close exit-4, text-only card) plus fake image determinism.
- Registered as the opt-in new-work-e2e suite; added test:new-work-e2e.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
"When the harness can X, do Y" hands the model an exit before the
command arrives; the observed reviewer skip walked through exactly that
door. The three gated constructions now lead with the imperative,
present the decision visually, open the chosen card, spawn the finish
reviewer, and carry their fallbacks as trailing clauses for sessions
that genuinely lack the capability. Constructions that already led with
the command keep their routing clauses unchanged.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The eb686f36 session read the separate-reviewer rule and spawned
nothing: an unnamed "separate agent" is an improvisation prompt, not an
affordance. The skill now ships impeccable-finish-reviewer next to the
asset producer: persistence first, ceiling against the card and comp
second, contract promise by promise, truth; ordered material fixes
back to the parent, no editing, no second detector. new-work names it
so the finish step invokes a thing that exists.
The asset producer was gated providers: codex, so Claude Code never
shipped it; the gate is removed and its two codex-only workflow lines
made provider-neutral with codex blocks.
Dist rebuild still deferred for the running campaign.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
codex.md becomes visualize.md and loads for every harness with any
image generation, native or the API fallback: after the direction
locks, three distinct compositional comps are rendered and put before
the user for approval, in-harness when it can display images,
otherwise on the decision page. Three is the number; one comp invites
rubber-stamping, and this approval round has repeatedly produced the
most compositional and ambitious work, so new-work now marks it
never-skipped. The codex-only subagent stays as a codex note.
The recent rule additions are tightened by a third: the asset and
imagery bullets merge into one, the canon exit loses its restatements,
the DESIGN.md-rule and chosen-card and ceiling clauses each shed their
second clause saying the first clause again. Same laws, fewer words;
prose that grows without bound recreates the attention gravity it was
written to fight.
Dist rebuild still deferred; the release-gate campaign reads the
pinned dist.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Em-dashes are used legitimately by humans, so em-dash-overuse fired far too
often. Reclassify it as the first advisory-tier rule: detected, but never a
failure.
Engine
- Add `advisory: true` to the rule metadata schema (em-dash-overuse is the
first). findings.mjs stamps `advisory: true` on advisory findings so every
consumer can partition without a registry lookup. Rule count stays 58.
- Raise the firing threshold from a flat 5 dashes to two gates: an absolute
floor of 8 and a density of about one dash per 500 characters of body text.
A long article that uses a few em-dashes no longer trips; a short,
dash-per-clause page still does. Entity decoding (mdash, numeric, hex) is
unchanged. Thresholds live in shared/constants.mjs so every engine agrees.
Browser parity
- The browser bundle carried a registry entry but no logic, so the overlay and
extension could never flag it. Add checkEmDashOveruse / checkEmDashOveruseDOM
in rules/checks.mjs (reads rendered text, no entity decoding needed), wire it
into the injected page-level pass, and carry the advisory flag through
serializeFindings so the overlay/extension can render it with the mildest
affordance.
CLI
- Advisory findings print under a separate dimmed "Advisory" section, are
excluded from the failure count, and never change the exit code (an
advisory-only scan exits 0). JSON keeps them with `"advisory": true`.
`--no-advisory` suppresses them entirely.
Hook
- Advisory rules are skipped by default in both the per-edit and Stop deep-pass
hooks, so the hook never nags about them. Opt in with
`.impeccable/config.json` -> `detector.advisoryRules: "include"`.
Tests
- Fixture + threshold + browser-adapter coverage; advisory-skip default and
opt-in for the hook; formatFindings partitioning. The em-dash-overuse stand
for a deferred copy rule in the tier tests is swapped to marketing-buzzword.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
From Paul's approved UX and the eb686f36 session post-mortem:
The standing exit: direction rounds carry a quiet, permanent "Play it
straight" action (payload flag canon, reserved id) on the decision page
and as the last structured-tool option. It is the user's door, never
the model's: never recommended, never weighed against the roll, and
choosing it swaps the bar rather than lowering it, two or three named
reference products become the craft level, canon executed at full
commitment. Safer/conventional steers resolve here, never to a
stranger re-roll.
Session fixes, each mechanical where possible: the ANSWER line now
names the chosen card's hero and board and directs opening them before
code (the session built from text alone after viewing a different
world's card); generation scale joins the imagery rule (a library of
centered 128px subjects foreclosed the atmospheric hero); DESIGN.md
rules are checked against the world's native devices and never added
to silence a hook finding (the session banned arcade lettering's own
offset shadow and laundered 8px through the ramp); staging joins the
FORM contract block (the axis was dropped silently at world-choice);
the finishing reviewer audits the ceiling against the QUALITY BAR card
after persistence (floor rigor was disguising unreached ambition); the
icon-tile clause names hand-drawn icons as remedy, not target.
Dist rebuild deferred: the release-gate campaign reads the pinned dist.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The structured-tool channel collapsed to a single direction plus
re-roll, which read as "the system only ever offers one idea" next to
the multi-card decision page. Both channels now share one structure,
assigned direction leading, the one or two fused challengers that
survived the weighing as named alternates, re-roll with steer, and
differ only in richness. The anti-lineup rule stays precise: what never
appears is a ranked menu of the model's own grounded candidates; dealt
challengers carry no ranking rut.
Note: dist rebuild deliberately deferred; the release-gate campaign is
running against the pinned dist and rebuilding mid-run aborts it.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Abort signals do not cancel the TCP connect phase, so an unreachable API
stalled the seed ~10s before degrading. All API calls now share one
deadline, the roll fetch races it, and the CLI exits explicitly.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A grounded direction with no hero rendered a blank 16:9 void where the
card image belongs (seen live: the assigned Xerox Zine card led the
hand as a black hole next to two rendered challengers). An option with
no imagery now drops the media region entirely and leads with its
kicker and text; an option with only a board shows the board as its
front image with no flip.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
With the PRODUCT.md skip fixed, the Opus smoke unmasked the adjacent
gap: the model builds a new world and never writes DESIGN.md (zero
attempts), so the worker's requiresDesign assertion correctly fails the
run. Same disease, same treatment: DESIGN.md is now part of recording
the decision, written before the first build edit in the same stretch
as the direction contract, and the finishing reviewer checks
persistence first, before any craft point is scored.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The truth split already permits full-fidelity demonstration data, but
permission at selection time was not holding at build time: models that
would not author covers, names, or thumbnails compensated with chrome,
which is the content-starved look the detector hunts. Two build rules
make it a mandate: every blank the ask round left open is authored at
production fidelity (content is authorable, claims are labelable,
nothing is omittable; unanswered commercial claims ship as marked
placeholders with a replacement list), and when image generation is
available, generating the build's imagery is part of building rather
than a nicety.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A live retest showed the model dropping to the structured question tool
when the roll degraded: with no challengers and no cards it judged the
page pointless and presented one option in plain text. The degraded seed
output and the new-work rule now both state that degradation changes the
cards, not the channel; a browser session presents the assigned
direction as a single text-only card with re-roll on the decision page.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A traced Claude Code injection asserts for whole model families that the
user is not watching and cannot answer questions; it ships default-on
with no off switch, and it suppressed every interactive step of a live
run (interview skipped, PRODUCT.md inferred, decision page never
served). Prose in a reference file loses that argument, placement wins
it: context.mjs now emits AUTONOMY_DIRECTIVE_CHECK as tool-result
content in the working turn, telling the model such a claim is a
harness default, never session evidence, and to probe once with the
question tool before inferring. init.md makes the same test mechanical:
tool presence proves an answer mechanism, one real probe round is
required, inference afterward must be labeled and disclosed in the
first reply. The degraded concept-seed path now also tells the model to
disclose the degraded roll instead of presenting it as a full one.
Image-gen signaling stays positive-only per Paul: key present emits the
capability, absence stays silent so harness-native tools are not
suppressed.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul reproduced the Opus smoke failure in a fresh repo: given a
natural-language build intent, the model runs concept-seed directly and
skips the init divert entirely, so PRODUCT.md never exists and nothing
grounds the challenger fusion. Prose already says init-first in both
SKILL.md routing and new-work.md; prose alone does not hold the floor.
The deal path now refuses with a NO_PRODUCT_MD directive routing to
reference/init.md when loadContext finds no PRODUCT.md. The --chosen
telemetry ping stays ungated.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The new self-detection ran before mode dispatch, so it also caught --wait,
--stop, and --schema. CI failed on the start/wait cycle test: --wait
returned 2 (no browser) where the documented poll loop expects 3
(WAITING). Under CI=1 the suite went 4 pass / 2 fail; it is 6 / 0 now.
Two of those modes were user-facing bugs, not just test breakage. --stop
exited 2 without killing the daemon it was asked to kill, leaking a
server process (verified: one daemon running, CI=1 --stop, still one).
--schema only prints a payload example, and new-work.md tells the agent
to read it before building a payload.
Detection can only tell whether this process can auto-open a browser, not
whether the user has one: SSH with a forwarded port and a harness with an
in-app browser both have a browser and no DISPLAY. The file already
treats serve-without-opening as first class, since --start spawns its own
daemon with --no-open. So the check now gates acquiring a session, not
managing or ending one. The blocking serve path still exits 2 on a
headless box, with a test pinning that.
Co-Authored-By: Claude <noreply@anthropic.com>
v4 changed PRODUCT.md's shape and retired the register axis, so an
upgraded project can carry answers nothing reads. Nothing measured that.
Two tiers, and the split is a performance contract:
- Boot (context.mjs, emitting CONTEXT_STALE) spends only what a boot
already spends: markdown already in memory, a bounded set of stats,
the small JSON files the boot reads anyway. No new directory walks.
One directive for the whole set, throttled to once a week per project
so a finding the user declined does not reappear tomorrow.
- doctor.mjs runs the deep pass on demand: git drift, ignore lists
validated against the live rule registry, hook script paths that stop
resolving, and the monorepo workspace sweep. --fix applies only the
migrations that carry no decision.
Findings are data, not prose, so the boot directive, the text report and
--json all render one set. Severity says what should happen: auto (fix on
the next write anyway), mention (state once), route (name the command
that owns the repair).
PRODUCT.md now carries a schema stamp so the checks stop reconstructing a
file's vintage from which sections it happens to have. Schema version,
not release version: a record written by 4.0.0 is not stale under 4.0.1.
DESIGN.md gets no stamp, because it follows the external design.md spec
that Stitch lints and every DESIGN.md signal is measurable without one.
The highest-value catch is a project that resolves to web while carrying
native build files, including a monorepo app inheriting a root record
that says web. That one costs output quality silently; nothing failed
before.
doctor follows the hooks/pin pattern rather than the Commands table, so
it stays out of the design menu and the count stays at 23.
Also corrects CLAUDE.md, which still documented the register axis,
reference/brand.md, reference/product.md, eleven deleted domain reference
files, and an extractRegister() whose only occurrence in the repo was
that sentence.
Prepared with AI assistance (Claude Code).
Co-Authored-By: Claude <noreply@anthropic.com>
The env-var bypass (IMPECCABLE_QUESTION_DISABLED) relied on the harness
remembering to set it. The script now also self-detects CI, SSH-without-
display, and displayless Linux and exits 2 with the structured-question
advice; --no-open skips detection (caller opens the URL itself, as the
tests do) and IMPECCABLE_QUESTION_FORCE=1 overrides it. new-work.md now
frames the decision-page rule by capability: open a browser if you can,
structured question tool if you cannot, exit 2 means fallback not error.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Setup step 1 tells the agent to follow context.mjs's directives, and the
no-PRODUCT.md-with-existing-code directive explicitly permits a narrow
refinement to proceed on the incumbent implementation and offer init
afterward. Routing's "Otherwise" branch said missing PRODUCT.md routes
through init, with no carve-out, so the two instructions disagreed on the
same request and the agent could block work context.mjs had cleared.
Rule 3 now splits the way the directive does: a new surface or
replacement world goes through init then new-work, a narrow refinement
proceeds and offers init afterward. Explicit and implied commands were
never affected; they route one rule earlier, which is what skill-behavior
scenario 10 already covers.
Co-Authored-By: Claude <noreply@anthropic.com>
Bugbot flagged the "resume without rerunning context.mjs" instruction
after init. It is right, and the gap is wider than the platform half it
named: context.mjs has two output branches, and the no-PRODUCT.md branch
omits DESIGN.md, the native platform references, and the unrecognized
`## Platform` warning. Because the skill never reruns the script once
init writes PRODUCT.md, whatever that first run withheld is gone for the
whole session. A greenfield iOS project would be designed without
reference/ios.md ever loading, and a project carrying DESIGN.md without
PRODUCT.md never saw its own design system.
The two halves need different fixes. DESIGN.md is authority in its own
right and does not depend on PRODUCT.md existing, so context.mjs now
emits it on both branches. Platform is unknowable before PRODUCT.md
exists, so no change to the script can recover it; init.md, the one step
that learns the answer, now loads ios.md / android.md / both right after
recording a native platform, and SKILL.src.md says so where it tells the
agent not to rerun.
Verified end to end against a temp project on both branches.
Co-Authored-By: Claude <noreply@anthropic.com>
Two true positives from the Bugbot review on PR #397.
document.md seed mode told the agent to run "Select one direction" for
paths A, D, or E. new-work.md has neither that heading nor the A/D/E
lettering since the workshop was restructured into named subsections, so
a literal read could skip the world-and-surface flow entirely. Point at
"Create or replace the visual world" and "Commit the world" instead.
critique.md let the heuristic table renormalize to an applicable maximum
when heuristics are scored n/a, but the report template hardcoded ??/40,
the rating bands only mapped raw numbers out of 40, and the persisted
meta carried total_score with no denominator. Trends could silently
compare 24/32 against 30/40 as if they were the same scale. The template
now prints the applicable max, the bands fall back to percentages for
partial sets, the snapshot records max_score and na_heuristics, and the
trend line states its denominator or breaks it out per run when they
disagree. critique-storage.mjs serializes frontmatter key-agnostically,
so the new keys need no code change.
Co-Authored-By: Claude <noreply@anthropic.com>
IMPECCABLE_CARD_BASE overrides the quality-bar URL prefix so eval
workers serve cards from the local checkout while impeccable.style
stays undeployed. IMPECCABLE_QUESTION_DISABLED makes serve-question
exit immediately with the structured-question fallback line, so
headless workers never block on a browser page nobody will answer.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Gate2 measured it precisely: on matched assignments Opus renders warm,
bookish, and child-facing subjects as cream, serif-italic, and
lamplight while Sol renders the same positions saturated, and neutral
prose hardening did not move it. The codex and gemini blocks set the
precedent for provider-conditional counterweights; this adds the claude
block at the palette decision: the first palette is already spent, an
OWN-WORLD block reading cream/paper/parchment/lamplight for an unpinned
Persuade surface is a failed rendition to rework from the world's
saturated materials, and nothing about the subject requires the
default. Verified present in the claude-code dist and absent from
codex.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Emit .grok skills, agents, and PostToolUse/Stop hooks; wire the CLI
installer and downloads; fix the plugin install path to #plugin; and
document Grok in HARNESSES.md and README.
AI assistance: written with Grok Build.
Paul's question exposed the blind spot: a closed tab left the agent
waiting out the full timeout. The page now sends a heartbeat every five
seconds while open; the server stamps lastBeat into the state file, and
--wait reports PAGE CLOSED with exit 4 when the beats stop for fifteen
seconds without an answer. The prose defines the fallback ladder:
re-present once through the structured question tool, then proceed
unattended with the assigned direction, stating assumptions.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The loading hand now has the true proportions: 16:9 shimmer media, a
tier line, a title line, three detail lines, and a button-shaped block
pinned to the card bottom, with each skeleton inheriting the measured
height of the card it replaces.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Live-fire bug from the demo: collecting a re-roll answer deleted the
state file, so --update could not find the still-running server and the
next hand had nowhere to land. Cleanup is now terminal-only: a re-roll
consumes just the answer file and leaves the server state for --update;
any other choice cleans up fully as before.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Two upgrades from the live session. Re-roll no longer ends the page: in
detached mode the server stays alive, the client gathers the cards back
into the center stack, deals skeleton cards with the site's shimmer,
and polls /next-status; the new --update mode delivers the next hand
and the page reloads into the fresh deal. Choices other than re-roll
still resolve and exit as before.
And the headline bug had a root cause: the fonts link never loaded the
weights in use, so the browser synthesized a fake 300 that read
off-brand and muddy. The link now loads Alumni 100 and 400 exactly, and
the h1 wears the homepage hero display role: weight 100 at
clamp(2.6rem, 5vw, 4.2rem), champagne.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
#lightbox { display: flex } outspecified the UA's [hidden] rule, so an
invisible full-viewport layer sat over the page and blocked all hover
and click. #lightbox[hidden] { display: none } restores reality.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Craft-gate forensics on matched hands: both models obeyed the same
assigned indices, but Opus rendered every kids cell as cream paper,
lamplight, and serif, and sample 1 chose Fraunces off the reflex list
with a bookshop-signage rationalization, while Sol rendered the same
positions as indigo bookcloth, coral thread, tomato, and marigold. The
dice work; the rendition prior escaped through two hatches, now closed:
naming a reflex face requires a reason no other face satisfies and a
subject association is never that reason; and bookish or child-facing
subjects do not soften the calibration, because cream paper is the
smallest corner of the book world.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul's motion and exploration pass. The reveal is now a real deal: the
cards begin piled at the grid center, blurred and slightly rotated, and
travel to their seats with a 110ms stagger on the brand ease (JS
measures each card's seat, so the pile works at any grid shape;
reduced motion skips it). Hovering a card bleeds its hero into the page
ground behind a lacquer scrim. An expand chip beside Board opens
whichever face is showing in a zoom-out lightbox with Escape to close.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
No cropping and no pillarboxing: the board spans the card width at its
native 16:9 exactly like the hero on the front, with the deep-lacquer
ground below and the label-and-CTA bar pinned to the card's bottom edge.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The design-system board is an information sheet; the back face now
contains it fully on a deep-lacquer ground rather than cover-cropping
its top and bottom.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul's read: flipping only the image inside a static frame looked
unconvincing. The card is now the object: front face carries hero,
lineage, title, and CTA; the back face is the design-system board at
full card height with a slim bar keeping the label and Build-this
reachable. The outer card keeps fan, deal, and hover; each face carries
its own lacquer chrome and the rolled card's gold ring rides both
faces. 700ms preserve-3d turn, instant under reduced motion.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
