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>
"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>
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>
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>
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>
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>
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>
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 call: start mode never auto-opens; the agent is alive and opens
the printed URL itself, in-app browser first, then the system opener,
then showing the URL (--open forces the system browser from the script).
The prose now leads with the start/open/wait flow and keeps the blocking
auto-open path for harnesses that can background a shell.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul's two concerns with the blocking design. --schema prints the exact
payload example so the model never guesses the shape (new-work.md points
at it). And harnesses that cannot leave a shell blocked (or cannot open
a browser while blocked) get a two-phase path: --start daemonizes the
server and returns the URL plus a key immediately, --wait polls for the
answer with exit 3 meaning ask again, exit 2 meaning the server died,
and --stop for cleanup. The browser open happens from the detached
server process, so it works even when the agent thread is short-lived.
State lives under .impeccable/questions/.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul's design, three pieces:
serve-question.mjs: the world decision presented as a themed page instead
of a text prompt. The script serves an impeccable-styled option board
(assigned direction leading with THE ROLL badge, dealt challengers as
alternates carrying their QUALITY BAR cards, re-roll and steer built in),
prints the URL, opens the browser, and blocks until the user chooses;
the answer lands on stdout as ANSWER JSON, so the shell call itself is
the wait and no harness machinery is needed. Local images are served by
the ephemeral server; nothing leaves the machine.
generate-image.mjs + context.mjs IMAGE_GEN_AVAILABLE: when an OpenAI key
is in the environment, context reports that image generation works even
without a harness-native tool (gpt-image-2, billed to the user's key,
stated before first use; Google skipped by decision). Harness-native
tools always win when present.
new-work.md: visualize-before-build is now the default whenever any
image generation exists, not a codex.md special case; the attended
presentation prefers the visual decision page and falls back to the
structured question tool. Evals keep the unattended path untouched.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul confirmed the craft-bar experiment: builds that saw the dealt
worlds' hero cards produced visibly stronger execution than the
no-image control. One clause makes the mechanism reachable for
harnesses that read only local images: download the card, then view it.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul's directive: the rendered board and hero for each dealt world ride
along with the challengers, framed as a quality bar (the finish and
commitment level the build is expected to reach), never as a mockup to
copy. The seed prints QUALITY BAR urls per challenger, preferring
API-provided cardBoard/cardHero fields and deriving from the concept id
otherwise; new-work.md instructs image-capable harnesses to view them
for the world being built. Server side, the roll API now returns
cardBoard/cardHero per challenger (impeccable-site).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Discussion outcome with Paul: a mediocre material world loses to
excellent abstract craft, but an unanchored "just be beautiful" escape
hatch would hand selection straight back to the model's priors. The
resolution: abstraction enters as named systems with their own grammar.
The derivation now states that the audience's graphic and screen
traditions (notation, publications, identity programs, data graphics,
interfaces) are as concrete a candidate as any physical artifact. The
catalog side of the same decision is a 12-entry abstract-graphic
authoring round in impeccable-site, pending review.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Smoke findings (Paul's review): the kids-reading derivation produced
seven candidates from one material family despite the divergence line,
and the brief-pinned bookshop world was rendered as the generic AI
bookshop (cream, serif italic, soft glow). The list must now span at
least three material families, and a pinned world licenses its full
material range, never just its softest rendition.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul's probe review (recovery-ab): the no-invention rule was blocking
bold greenfield directions whose demonstration data does not exist yet;
kids-reading amplified the product's "quiet support" adjectives into a
whole-page aesthetic; and nothing guaranteed a Persuade surface still
sells once the form commits (a prior generation shipped zero nav and
zero CTA in the first viewport).
- Truth split in two: commercial and factual claims stay uninventable;
illustrative material is authored at full fidelity, labeled synthetic,
with a replace-with-real list for the user. Mirrored in the build
section so execution-only sessions get it too.
- Persuade floor restored from a22: conversion lives inside the form's
own vocabulary (one-line hook, visible primary action, legible reading
order); a committed form that hides the offer has not finished
translating. The contract's FIRST VIEWPORT block now names where the
primary action sits, and the finishing review verifies the mode did
its job.
- Calibration: negative constraints rule out devices, not exuberance;
product-behavior adjectives do not dictate surface energy.
- Web leverage: when the chosen world names a technique (canvas, WebGL,
view transitions), build the technique, not a static imitation.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Probe finding (recovery-ab, obs sample 1): "treat both structures as the
rut, not the range" let the model rank the observability dashboard grid
at position 7, and the dice landed on it; the challenger fusion rescued
that draw, but a die face spent on the category's own page is a wasted
roll. The a26 wording excluded both structures outright; restore that
with the reason attached.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The ship40 concept pipeline had reversed the proven a-series mechanisms:
the seed's roll decayed into a shortlist nomination that taste functions
(model ranking, candidate floor, simulated user) then argmaxed into the
safest card; the costume check returned as the Translation veto and
carrier-removal test; and the 07-15 rewrite deleted the calibration,
reflex-font lanes, color strategies, and commit-every-atom language that
had held off the cream-editorial default since the alpha era. Five of six
frozen craft directions converged on the same warm-paper family and both
builders obeyed them.
This lands the repair on top of the in-progress simplification:
- new-work.md: the script assigns the build index again on both scopes;
catalog challengers are fused (challenger supplies form and grammar,
product supplies every fact, clarity wins conflicts) and weighed on the
two proven axes only; attended runs present one fully committed
direction with re-roll and an optional steer instead of a ranked
lineup; the color-strategy picker, reflex-face list, saturated-look
calibration, first-viewport thesis and memory test, commit-every-atom,
scroll pacing, and prove-don't-claim return; the direction contract
returns as five lean blocks audited by the separate-agent finish.
- concept-seed.mjs: PROMOTED INDEX becomes ASSIGNED INDEX with
build-assignment semantics; self re-roll only on named factual grounds.
- craft-floor.md: hook-active sessions act on findings instead of
re-auditing; the Refuse list is framed as category defaults the brief
can earn; a closing commitment line keeps a ban list from being the
last word before code.
- codex.md / shape.md: contract references restored for flow coherence.
Adopts the concurrent session's ceremony cuts, softened challenger
instruction, seed SOURCE IDs and --candidate-count, detector-ownership
fix, and the removal of the hook-side contract audit (the audit now
belongs to the separate reviewer at finish).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Replace the Consequence floor with Signature: one authored move that
makes the experience unmistakable and shapes implementation, named in
terms of what the visitor experiences. Translation now demands the
source's aesthetic and compositional laws survive alongside product
structure, so function-without-character reads as safe flattening and
character-without-structure as costume.
Add an expand-then-contract step before the direction contract: decide
spatial, motion, interaction, narrative, and system questions as one
studio plan that causes itself, then compress into the contract. Staging
guidance follows the seed's move to several inputs.
Co-Authored-By: Claude <noreply@anthropic.com>
/api/roll deals deterministic challenger rolls server-side (same salts and
sha256 ranking as the local seed, verified bit-for-bit); the request log is
the impression record. /api/chosen takes the anonymous choice ping. Events
land in Workers Analytics Engine.
concept-seed.mjs resolves data in order: local catalog dir, roll API,
degraded promotion-only seed. --chosen sends the choice ping; DO_NOT_TRACK
and IMPECCABLE_NO_TELEMETRY disable it. API-dealt seeds carry the telemetry
instruction inline.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Opus probe (r10-opus-wireframe): control articulates loose skeletons,
wireframe arm names them (dubbing script sheet, timecode gutter spine)
and diffs against the standard stack every time. Targets Paul's
layout-diversity question: concept-atom commitment with template
skeletons underneath.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul: 'way better to have the first iteration land fully committed to
the concept, because that's the genuinely hard part. the next pass can
make sure it is clear and effective.' The check selected against the
original lektor site itself, the campaign's 10/10 reference.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The inverse-probe extracted this from Paul's NewRelic review and it won
in the batch4 forward test; it was never ported. Craft diagnosis: pages
lose on rhythm monotony (one treatment uniformly applied), not defects.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul: no stochastic challenger-assignment mode (unreproducible bad draws
= undebuggable bug reports); keep the weigh-off. Every roll now prints
its key so any field report can be replayed with --from.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Contract-probe campaign findings (evals repo, notes/fable-oneshot-craft-plan.md):
a single model's resonance ranking is deterministic (30/35 identical
concepts across 16 framings); dice must come from the script, mirroring
the palette-seed result. Derived candidates stay grounded in the
audience's world + subject's cultural home; challengers win only on
identification x clarity; incumbent-with-deliberate-idea overrides the
roll. Validated at contract level on 01-observability + r10-lektor
(teletext ranks #3 for lektor; assigned index 3 produced it).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Four lines from the r10 dual consultation (codex gpt-5.6-sol + gemini
3.5-pro on the actual HTMLs) and the hero-probe micro-eval: the probe
isolated a first-viewport monoculture (same split template in every
sample, control and skill alike) and showed these lines break it while
codex's raw 15-liner alone does not. The incumbent sentence swap fixes
the r10 root cause both consultants independently identified.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Transcript evidence (a12 01-observability): plans commit and deliver on
the axes with contract-strength language (palette, type, even theme
inversion) and stay default on the axis without one (layout gets a
single conventional breath). And plans living in invisible reasoning
means nothing can hold a build to its intent. The direction is now
written as a comment block at the top of the artifact answering: the
concept, the hour-later memory, why not the modal competitor page, the
signature, the first viewport's move. Critics and evals can score
delivery-against-contract; a mood is not an answer.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul's a10 review: palettes are refreshed (the palette-exclusivity
line's fingerprint) while layouts stay boring in every version. Same
cure, same shape: the layout has exactly two legitimate sources, the
concept or the content's own structure; the category's habitual
skeleton is neither.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Paul's a11 review: heroes are safe SaaS viewports, everything
predictable; mobile Operate ships dark despite a brief that specifies
outdoors-in-motion use. Decide-then-build now opens with three
one-line directions differing in concept (the instinctive pick that
any studio would reach for is the default wearing your name); the
Operate mode adds: the usage scene is part of the spec, the theme
follows the scene, not the category's habit.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Per Paul: rather than gating a second file, fold what made the craft
path superior into the file both models already read 21/21 through the
gate. new-work.md gains 'Decide, then build' (direction as one
confirmable paragraph; attended pauses, unattended records-and-goes;
codex.md mock flow when image generation exists) and 'Finish like a
studio' (inspect, honest critique, patch, detector). craft becomes a
deprecated alias like teach: invoking it forces attended checkpoints,
nothing else differs; the reference is a redirect stub. codex.md
retargeted. Existing-world feature builds remain governed by the core
floor (unmeasured path, noted in the plan doc).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
a7 transcript evidence: 01-observability samples drew orange-honey and
green seeds, recited the color-strategy menu, and shipped dark
category-reflex palettes anyway; the model applied the subject's
workmanlike grammar to its own landing page. Two generic lines: the
mode belongs to the surface, not the subject (a landing page for a
dense tool is still Persuade; deciding a page can be plain because its
subject is workmanlike is the category error in reverse), and the
palette has exactly two legitimate sources (seed or the subject's
world; the category's habitual palette is neither).
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
