A codex field run dealt six challengers into an eight-sketch batch
behind an opaque subagent, and the user stared at a page of shimmer
asking whether anything was happening at all. Four fixes from that run.
A hand now holds at most three challengers, the rest banked for
re-rolls, so fairness within the hand stops multiplying into a queue.
Sketches greek everything but the product's real name and one real
headline, because an invented spec, price, or ship date in a sketch is
a claim PRODUCT.md never made, and comps have solved this for a century.
Sketch production follows the user's reading order with the first file
doubling as the producer's heartbeat, and the parent's --wait loop
checks the sketch directory each pass, reclaiming the batch inline when
two minutes pass with nothing landed. And a failed --start now captures
the daemon's stderr to a per-key log and names the sandbox as the usual
suspect, instead of reporting only that failure occurred. The shimmer
counts its elapsed seconds, and gives up at 150 instead of 300.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Three field notes from a live review. The deck now escapes the content
column and runs edge to edge, so a cut-off card sits at the screen edge
where it reads as more cards instead of at an invisible container edge
where it reads as a bug; the first card still aligns with the column
via scroll padding. Whichever side hides more content wears a fade, and
a hard edge means the end. The vertical pager grows from a bare chevron
into labeled Back and More pills, because in a column deck it is the
primary way forward. And hovering the inspiration thumb now takes over
the whole media region instead of a timid zoom; the sketch is the
promise, the inspiration is a glance, and the glance must cost nothing.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Field-checked in a real browser, which surfaced three defects the DOM
tests could not: the generic .media img display rule defeated [hidden]
and floated an empty block over the shimmer and its sketching note; the
deal animation left every card at opacity zero in an unfocused tab,
because rAF throttling is real and decoration must never gate content;
and the sketch poll's cache-busting query missed the anchored /img
route, so a landed sketch kept shimmering forever.
The grid is now a snap-scrolling deck: one row in a wide viewport, one
column in a tall one, with edge arrows that appear only on overflow and
page one card at a time.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
The decision page compared unlike things: the grounded direction was a
wall of text beside curated catalog art, the catalog art read as a
promise of the build, the weighing silently shrank the challenger set,
and the standing exit hid in the footer under the cards it must not
soften. Every card now shares one anatomy (thesis, palette chips,
material tags, first viewport, case, risk), every dealt challenger is
presented with the weighing written on it rather than applied to it,
the catalog image rides picture-in-picture as labeled inspiration with
the lightbox a click away, and canonCard renders the category standard
as one honest, subordinate card.
When image generation exists, each card declares a sketch slot the page
polls: serve first, generate after, through one shared deliberately
unfinished frame, so the comparison stays about direction instead of
rendering luck. The asset producer takes the batch when subagents
exist; the chosen sketch returns in ANSWER to seed at most one comp
probe, and the comp round still renders its full set.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Raising the reviewer's turn budget did not change its fate, only its
reading: 43 tool uses instead of 22, still reaped mid-read with nothing
written, because the SDK ends a run at max-turns without warning and the
model never feels the deadline. The definitions now carry the deadline
themselves: reading is an allowance, batch Reads per turn, take the
decisive inputs first, sample instead of walking the tree, and write by
mid-budget, naming what went unread. A review built from what you saw
beats a perfect review that never arrives.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A codex run turned an approved comp into a related second art direction
and the finish reviewer passed it: the review anchored on the direction
contract, a lossy abstraction the builder wrote, and every element that
abstraction dropped passed silently. Four changes close that chain. The
reviewer inventories the comp's salient elements before reading the
contract and classifies each one (match, adaptation, missing,
contradicted, added without approval), with adaptations citing the
answer, brief, accessibility need, or product truth that forced them,
and fidelity failures outranking craft in material_fixes. The visualize
inventory gate records compositional commitments alongside asset media,
since the 150-word contract cannot carry them. The north-star allowance
now says what it permits: translation, never recomposition. And the
finish sequence recaptures the same viewports once after the fix batch,
so what the documenter records is what actually shipped.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A finish review reads the artifact, two full-page screenshots, the
approved comp, the quality-bar cards, and the contract before it may
write a word; at max-turns 12 the SDK reaps it mid-read and the parent
receives the opening sentence as the whole review. Observed twice in a
row (spawn and respawn) on the first real subagent run. The documenter
reads at least as much, and the asset producer pays per asset.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Round eight closes the stale-local class completely: the develop
candidate sits before the remote-default entries, so when origin/HEAD
itself points at develop, its name claim let a stale local develop win
over the fresher origin/develop. The candidate now leads with any
remote-advertised develop rev, exactly as the remote-default and
upstream candidates already lead with theirs. main/master were already
covered since their remote-default entries come first in the order.
Failing-first test forces local develop two commits behind.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Round seven: a remote-advertised default candidate tried the local
branch first, so a stale local main outranked the fresher origin/main
the symref points at and refilled changedFiles with the divergence.
The candidate now leads with the advertised remote rev, mirroring the
upstream candidate's reasoning. Failing-first test: local main forced
two commits behind the remote default, feature delta stays clean.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Round six, and the upstream-parsing ambiguity dies at the root: @{u} is
now resolved via rev-parse --symbolic-full-name, where refs/heads/...
IS a local upstream and refs/remotes/<r>/... IS remote-tracking. The
previous remote-membership heuristic still misread a local feature/foo
upstream when a remote literally named "feature" existed. The
adversarial test now configures exactly that remote and passes.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Round-five bot findings, one real root cause: splitRemoteRef treated the
first slash in any ref as a remote separator. A local upstream named
release/2.0 was truncated to "2.0", and feature/foo tracking from branch
foo collapsed to the current branch's own name and was self-skipped,
discarding a valid base both times.
The split now happens only when the prefix names a configured remote;
otherwise the whole ref is one local branch name. The per-remote HEAD
symref loop strips its own queried prefix directly (that remote may be
fabricated in tests or partial clones without appearing in git remote).
The reported pruned-upstream shape already resolves via the multi-remote
rev lists from the previous round; its test now guards that.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Cursor and Greptile converged on one root cause from the previous round:
candidate revs stopped at origin (develop tried only develop and
origin/develop; a remote-default entry carried only its own rev), so the
name-level dedup discarded a same-name base living on another remote. A
fork-parent layout with develop only as upstream/develop, or a pruned
origin/main beside a live upstream/main, lost its base entirely.
revsFor(name) now expands to the local branch plus <remote>/<name> for
every remote (origin first), and all named candidates use it, which is
exactly what makes the dedup safe. Two failing-first tests cover the
upstream-only develop and the pruned-origin/live-upstream main shapes.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Two more real gaps from the post-rebase review round: a detached
checkout reads its branch as the literal HEAD, so the integration guard
never fired and candidate selection could diff a detached tip on main
against develop; and the remote-default check only consulted origin, so
a fork-parent layout whose only remote is upstream lost the guard on
its default branch entirely.
The guard now treats a detached HEAD as no-diff-base, and default-branch
symrefs are collected from every remote (origin first), feeding both the
guard and the candidate list. Two failing-first tests cover a detached
tip beside a diverged develop and an upstream-only trunk default.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
The probe asked whether node could load ESM, while the notice promised a
Node 22 floor and package.json engines declares >=22.12.0. Reviewers kept
flagging the gap, and they were right to: a 14.18-to-21 runtime passed the
probe on the strength of one import while the hook and its detector bundle
are only ever exercised on the engines floor, so "can load our code" was a
weaker claim than the one being made for it.
Check the floor directly: parseInt(process.versions.node) >= 22, in
ES5-only syntax that parses on any node old enough to fail it. Probe and
notice now derive from one NODE_MAJOR_FLOOR constant, so they cannot
disagree, and the archaeology about node: scheme support and pre-15
unhandled-rejection semantics goes with the import it explained.
Add the missing contract test: every generated hook command carries the
probe, the notice appears exactly where a harness can render it (Claude
and Codex, project and plugin), and the expected floor is read from
package.json engines rather than repeated by hand.
Verified against a fake pre-22 node, no node, and a real node: one notice
then the marker holds it silent, exit 0 in every failure shape, and the
hook's own exit code still passes through on a supported runtime.
Co-Authored-By: Claude Fable 5 (via Cursor) <noreply@anthropic.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Greptile flagged that the probe does not enforce the Node 22 engines floor.
Two parts to that, and they land differently.
The real defect is narrower and worse than stated: the hook closure imports
`node:fs`, `node:os`, `node:path` and `node:url`, and the `node:` scheme needs
14.18, so a bare `import('fs')` probe passed on 12 and 13 and those runtimes
then died on the real import, which is the banner this branch exists to remove.
Probing `node:fs` closes that. The added `.catch(()=>process.exit(1))` is load
bearing rather than tidiness: before Node 15 an unhandled rejection is only a
warning and the process still exits 0, so a rejected probe would have read as a
pass on exactly the versions in question.
Not enforcing 22 is deliberate and stays. The probe asks whether this runtime
can load our code, not whether it is a supported one, so a 14.18-to-21 runtime
that works today keeps working rather than being silently switched off. The
notice names 22 because that is the version worth installing, and it only ever
reaches someone whose runtime already failed the probe, so no user is shown a
threshold that contradicts what ran.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Commit 8397d532 took a reviewer's word that Codex expects hookSpecificOutput
and dropped its notice on that basis. Codex documents `systemMessage` for
PostToolUse and Stop as text shown as a warning in the UI or event stream,
the same field Claude Code reads, so the notice belongs there and the earlier
comment asserted something unverified.
Checked the rest against their own references while here. Cursor's preToolUse
output is permission-shaped and its user_message renders only when the action
is DENIED, so warning would mean blocking the edit. Grok treats PostToolUse
and Stop as passive events and ignores stdout outright. Copilot's contract is
unconfirmed. Those three keep the probe alone, which is a verified limit now
rather than an assumption.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Bugbot caught the Codex plugin builder still invoking node directly, and the
same reasoning covers GitHub Copilot and Grok Build: all three shipped the
exact failure this branch exists to stop, and sat visibly inconsistent with
their guarded siblings.
Route them through guardedNode with no notice, matching Codex and Cursor.
GitHub gains a second property from it: outside a git repository
`$(git rev-parse --show-toplevel)` expands to nothing, so the old command
handed node a path that could not exist and failed the turn. The file test
now short-circuits that to exit 0.
Every builder carries the probe; only the two Claude manifests carry the
notice, which is the only harness whose response shape is confirmed.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
`systemMessage` on stdout is a Claude Code contract. The shared guard was
emitting it for Codex and Cursor too, where what a harness does with stdout
it did not ask for is unconfirmed, and a Cursor preToolUse hook printing an
unexpected JSON object is the wrong thing to guess about.
Pass the notice in per harness instead of baking it into the guard. Claude
manifests opt in; Codex and Cursor take the runtime probe alone, so an
unsupported runtime stays as quiet there as it was before the probe existed.
Giving them their own shape later is one more argument at the call site.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The hook command invokes bare `node`. When that node predates ESM,
`hook.mjs` dies while it is still being parsed, before the script's own
always-exit-0 contract can run, so node exits 1 and the harness reports a
hook error on every Stop and every edit.
Probe the runtime in the command string before invoking the hook, and
route the Claude plugin manifest through the guard that already covered
the project-local manifests. On probe failure the command exits 0 and
emits a one-time `systemMessage` naming the two fixes available to the
user, since nothing written in ESM can report this condition.
Fixes#410.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The suite runs real git subprocesses in its scratch dir, and on Node 22
the recursive afterEach delete raced git's object writes: rmdir of
.git/objects threw ENOTEMPTY and failed an unrelated PR's CI run
(seen on the #418 rebase run, checkDesignDrift suite). rmSync's
maxRetries/retryDelay options exist for exactly these transient errors.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Cursor Bugbot: sitting on a non-standard default such as trunk (the
origin/HEAD target) still ran candidate selection, where develop or main
could win and produce an integration-vs-integration diff. The guard now
treats the remote default branch as an integration branch alongside the
conventional names. Failing-first test: on trunk with a develop branch
present, the scope stays the working tree.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Cursor Bugbot's remaining round-1 finding held for the current code
too: in a git-flow repo whose platform default was never flipped off
main, a feature branch without an upstream picked origin/HEAD's main
over the develop branch features actually merge to, dragging the
develop-vs-main divergence into scan targets. develop now sits between
the upstream signal and origin/HEAD in the candidate order; repos
without a develop branch are unaffected. Failing-first test covers the
exact shape (develop exists, origin/HEAD -> main).
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Cursor Bugbot round two, both real: an upstream or origin/HEAD naming a
DIFFERENT integration branch bypassed the conventional-name guard, so
sitting on develop with the remote default at main still produced the
integration-vs-integration divergence this detection exists to prevent.
And splitRemoteRef returned null for a slashless @{u}, silently dropping
local upstreams (branch.<x>.remote = ".").
Base detection is now skipped entirely on an integration branch: no
signal may override the working-tree scope there. A slashless upstream
resolves as its own name and rev. Two failing-first tests: origin/HEAD
pointing at main while sitting on develop, and a feature branch
tracking a local canary branch.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Both review bots found real gaps in the first pass: candidates were
verified as local branch names only, so an origin/HEAD target with no
local checkout fell through, and stripOrigin() dropped upstreams on any
remote not named origin (fork workflows tracking upstream/release).
Candidates now carry a display name plus the revs to try in order: the
upstream's remote rev wins outright (it tracks the actual merge target,
so it beats a possibly stale local branch of the same name), origin/HEAD
tries the local branch then the remote-tracking ref, and the
conventional names each try local then origin/<name>. git.base keeps
reporting the friendly branch name while the diff runs against whichever
rev resolved. Two new failing-first tests: remote-only default branch,
and an upstream on a remote named upstream with no local base branch.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
context-signals hardcoded ['main', 'master'] as diff-base candidates, so
repos integrating through develop (or any other branch) diffed against
the wrong base: git.changedFiles carried the entire divergence and
downstream commands scanned the wrong set (issue #302).
The base is now detected, most specific signal first: the branch's
configured upstream (@{u}; a branch pushed with -u tracks itself and is
skipped by the self-check), then the remote's default-branch symref
(origin/HEAD), then the conventional integration names including
develop. The conventional fallbacks are withheld when the current branch
is itself one of them, so sitting on main in a repo that also has
develop keeps the working-tree scope instead of diffing two integration
branches against each other.
Five tests (three failing-first): develop-based feature branch,
origin/HEAD detection with a non-standard default name, upstream
tracking, on-the-integration-branch fallback, and the
integration-vs-integration guard.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Cursor Bugbot caught a real hole: accept unlocks at the first variant,
so a late generation agent_done for the same session id could arrive
after Accept and close the awaited failure window early, reopening the
exact #384 gap. The SSE broadcast carries no sourceEventType, so only a
carbonize agent_done is provably accept-side; the release is now gated
on it. Copilot's wording point led somewhere real too: a carbonize-phase
failure raises the same error after the source WAS promoted, so the
toast now says "may not have been saved" and normalizes the server
message's terminal punctuation. Regression guard extended to pin both.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Accept is optimistic: POST /events acknowledging the intent schedules
cleanupAcceptedSession(), which nulls pendingAcceptedSession before
live-accept.mjs has run. When the accept later failed (missing markers,
preview error, receipt conflict, source_locked), the SSE 'error' guard
keyed on pendingAcceptedSession could no longer match its id, so the
tailored recovery never fired: the user got a generic error toast, the
session was gone, and nothing said the variant was never written
(issue #384, analysis by Cursor Bugbot on #381).
Following the issue's fix sketch, an awaitingAcceptResult id is set on
the optimistic success path and deliberately survives the teardown. The
'error' case matches it and tells the user plainly that the variant was
not saved and to pick + generate again (post-teardown the wrapper may
already be gone, so restoring CYCLING is not honestly possible). The
marker is released when the real accept result arrives (complete /
accept / post-accept agent_done) or when a new session supersedes it.
Regression guard covers the set-before-teardown ordering, the error
match, and cleanupAcceptedSession leaving the marker alone; the existing
source contract now also asserts handleGo clears it.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Both bots caught the same false report: the summarize step ran off the
initial drift flag, so the no-drift-after-rebuild exit still claimed a
commit landed on main. The commit step now records pushed=true/false in
its step output and the summary reads it.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
Copilot's two review points: the fifth attempt performed a full
reset + install + rebuild + 25s backoff that nothing would ever consume
before the job failed, and the retry message blamed "main advanced"
when the combined condition also fails on push errors (network, auth).
The loop now breaks before recovery on the final attempt, and both the
retry and terminal messages name the two possible causes.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
The sync workflow built once from the checked-out main and aborted when
a human commit landed during the ~30s build window (about 10% of runs
per the evidence in issue #388), leaving generated provider output
stale until the next unrelated push re-triggered it.
The commit step now loops up to five times: on a lost race it resets
hard to the fresh origin/main (source included), re-installs and
rebuilds, and pushes again with linear backoff. Every attempt therefore
builds from the main it will land on, which is the invariant the old
abort guard protected; the merge-base check stays inside the loop as
the pre-push verification. When the rebuilt output shows no drift (the
racing commit was another sync, or the new source produces identical
output) the step exits cleanly instead of committing an empty sync.
Validated by yaml-lint, bash -n, and a local three-repo simulation
(bare origin + worker + racer) confirming the lost race rebuilds
against the racer's source and lands matching output on attempt two.
Retry design proposed by @mktdgtbrz in #388; implemented from the
description with the no-drift early exit added.
Prepared with AI assistance (Claude Code), directed by @pbakaus.
Co-Authored-By: Claude Code <noreply@anthropic.com>
compileProviderBlocks only processes standalone-line blocks, so the
inline codex spans leaked literal tags into every provider's agent
output, degraded fallbacks included.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Harnesses with no subagent capability now run each role inline from the
same single source. The build emits reference/degraded/<role>.md for every
agent in skill/agents/ (role name is the agent name minus the impeccable-
prefix), stripping frontmatter and prepending the inline-substitution
preamble. These pass through the same provider-block compilation and
placeholder replacement as ordinary reference files, so <codex> blocks and
{{placeholders}} resolve per target, and they land in the committed harness
dirs on build:release like every reference file.
Repoint the three capability-first fallback sites in the prose at the
generated files: new-work.md reviewer and documenter fallbacks, and
visualize.md asset-producer fallback.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Probe attribution on Opus 5 showed the screenshot bound working (42
to 16) while the real burner ran free: five rounds of node -e
micro-edits, eight rebuilds, and inline defect hunts absorbed the
reviewer's and documenter's jobs until the turn cap killed the run
mid-hunt. The two-round ceiling now names scans, micro-edits, and
rebuilds; after the second round the build thread stops polishing and
ships the rest through the reviewer (one batched fix pass, one
rebuild, stop) and the documenter.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
github-actions[bot]
Paul BakausandClaude Fable 5
dependabot[bot]andGitHub
