Merge origin/main into fix/review-fix-range-rereview; stamp v1.91.226.0 (#3414)

Co-authored-by: capy-ai[bot] <230910855+capy-ai[bot]@users.noreply.github.com>

CHANGELOG.md

@@ -1,6 +1,6 @@
 # Changelog
 
-## [Unreleased]
+## [1.91.226.0] - 2026-10-26
 
 **/review can re-review just its own fixes after a fixing pass, under the same rules as /ship (#3339).** Before, every /review fix cycle reran the whole review over the whole branch, so a large branch could hit the three-cycle stop on findings in code the earlier passes had already read. Now the pass after an edited pass starts with `gstack-review-log --start review --fix-range` and its checklist, specialists and QA read only what changed since that pass. The Step 4.8 adversarial review still reads the whole branch on every pass, because its rows carry their own freshness. A narrowed pass is never the final review: when it makes no edits, /review runs one full pass before writing the final record, and a fix-range record is never CURRENT. Garry's "do it" on 2026-10-24 for a fix-range re-review with a final full pass covers it.
 
@@ -26,6 +26,84 @@
   - letting Step 5.0 read this invocation's records.
 - No threshold, budget, timeout, trial count or case kind changed. Size pins only went down.
 
+## [1.91.225.0] - 2026-10-26
+
+**/design-html builds landing pages around one visitor action and pages that look finished, without inventing proof; adapted from Elaya's design skills (GSTA-92).** Freeform /design-html asked for purpose, audience and feel, then built. It never asked for the one action the page exists for or where the traffic comes from, and nothing caught a page that shipped with invented customers and numbers, a dead `href="#"`, a scroll-listener reveal, `100vh`, or no description, favicon or working skip link. Now every run reads a craft section at Step 1, and a free checker counts the tells at Step 4. Adapted from Elaya's [ai-design-skills](https://github.com/elayadesign/ai-design-skills) (landing-page-design, design-to-motion) and [redesign-skill](https://github.com/elayadesign/redesign-skill), MIT, © 2026 Elaya. Thank you, Elaya.
+
+### Changed
+- **/design-html reads `sections/craft.md` on every run (Step 1).** The spec opens with `Surface: Persuade` or `Surface: other`. A Persuade page (landing, pricing, waitlist, campaign, comparison) asks at most three multiple-choice questions (primary action, traffic source, published or reference) and states the rest in one `Assumed:` line; picks one of four archetypes with a reason; keeps the hero budget (one proof signal only when real proof exists, never a metric row) and chooses the rest of the page from the archetype and objections; gives the primary action real invalid, pending, failure and success states. Every page: claims (testimonials, logos, results, prices, guarantees) come only from the user, and missing proof is left out or marked `[Needs proof: …]`; illustrative data looks real; a ship floor scoped to full documents (title, description, favicon, share tags on shareable pages, `og:image` only from a supplied URL, a skip link that reaches `<main>`, landmarks, alt text, inline validation, no dead links, `aria-current` on route navigation); scale snapping with readability floors and the nested-radius rule; balanced line breaks; motion rules for the one authored moment (content visible without script, `IntersectionObserver`, cheap properties only, no `scale(0)`, named curves, `100dvh`).
+- **Step 4** runs `bin/gstack-design-craft` after the Slop Gate, whether or not the detector is installed, with the gate's contract: one fix pass, one rerun, the rest accepted with a reason, once more when the refinement loop ends. The hand-off opens with a "Before this can ship" block: proof markers, production dependencies, exceptions, and every claim the check changed.
+- **/design-shotgun**: a landing, pricing or campaign screen's variants also differ in page archetype, keep the one primary action identical, and carry their section outline as a caption.
+
+### Added
+- **`bin/gstack-design-craft`** (`lib/design-craft.ts`): prints `CRAFT_CHECK: <id> <count> <fix>` per tell and `CRAFT_RESULT: clean` (exit 0) or `<n> issue(s)` (exit 1); exit 2 on a non-HTML, empty or unreadable file. Copy rules read decoded visible text; `alert()` counts only in scripts; head and navigation rules run only on a full document. Buzzwords report under the catalog's `marketing-buzzword` id.
+- **NOTICE.md** and `licenses/MIT-elaya.txt`. Not carried over, because they conflict with gstack's design rules: Elaya's fixed fonts (Geist and Poppins are on gstack's overused-display list), palette, spacing values and icon sets, gradient hero text, mandatory scroll reveals and tagline section, and the island nav.
+
+### Tests
+- `test/design-craft.test.ts`: every tell with its count; a finished page, inlined Pretext, a `100vh` line with its `100dvh` fallback, discount copy and in-page anchors stay clean; look-alike attributes, encoded markers, `!important` and reversed fallbacks; skip links judged by position and target in any language; the craft section's banned words and names all flagged; CLI exit codes.
+- `design-html-slop-gate` also requires the craft check to run once or twice; the design-html carve guard requires the craft section read.
+
+## [1.91.224.0] - 2026-10-25
+
+**`/context-restore` now says when commits landed after a checkpoint was saved, instead of offering finished steps as the plan.** `/context-save` recorded the branch, a timestamp and the modified files, but not the commit, so a checkpoint saved before a step was committed came back with that step still listed under Remaining Work. Ported from #3316 by @soyames, with their commit kept.
+
+### Fixed
+- **`bin/gstack-slug --stamp-checkpoint`** also writes `head:`, the commit the save describes, next to `remote:`, `project_root:` and `worktree:`. A re-stamp replaces it rather than adding a second line.
+- **`/context-restore`** compares `head:` with `HEAD` before presenting. When commits landed after the save it prints how many and lists up to ten, and puts every Remaining Work item under Verify first. A saved commit that is not in this checkout's history (saved on another branch, rebased, reset, or not fetched) is reported the same way. A malformed `head:` is checked as 40 or 64 hex digits before any git command sees it, `head:` is read from the frontmatter only, and outside git the comparison uses the saved `worktree:` before `project_root:`. Checkpoints saved before this keep their usual sorting with a one-line note.
+
+### Tests
+- `test/context-restore-head.test.ts` runs the real save-path, stamp and head-check blocks from the generated SKILL.md files in fixture repos: unchanged, two commits landed, diverged after a reset, resumed from another branch, a nested linked worktree restored from outside git, a re-stamp after a new commit, pre-tracking, a `head:` that appears only in the body, and a `head: $(touch pwned)` that must never execute.
+
+## [1.91.223.0] - 2026-10-25
+
+**/plan-tune's inferred profile finally moves: a clicked answer is logged as the question's registered option key, not its label.** #2623 showed every inferred dimension stuck at 0.5 on a 1,286-event log: the AskUserQuestion hook recorded labels like "A) Fix it now (recommended)", while the signal map that turns answers into dimension nudges is keyed by registered option keys like `fix-now`, so 0 of 1,286 answers matched. Now five hook-captured label answers to the registered architecture-finding question move `architecture_care` from 0.5 to 0.632, and `--trace` lists all five.
+
+### Fixed
+- **`bin/gstack-question-log`**: on a hook-captured event (`source: hook`) for a registered `question_id`, the clicked label is mapped to that question's registered option key before it is written, and the label is kept in a new `choice_label` field. A label in `recommended` is mapped the same way, so `followed_recommendation` compares keys. Ad-hoc and `hook-<hash>` ids and unmatched or ambiguous labels are written as given. Rows from other writers (the agent's own log call, auto-decide, Codex import, typed free text) are unchanged, because readers such as the plan-ceo-review auto-decide check match the full mode names those rows carry. Existing logs are not rewritten.
+- **`scripts/psychographic-signals.ts`**: new `resolveOptionKey(question_id, label)`. It strips the `A)` selector and `(recommended)` suffix, then matches the label's leading words against the question's own options only: each key's own words (`fix-now` → "fix now") plus a short cue list ("Add to TODOs" → `defer`, "Ship as-is" → `skip`, "HOLD SCOPE" → `hold`). The longest cue wins and a tie stays unmapped.
+
+### Tests
+- `test/plan-tune.test.ts`: every registered option resolves from its own words, bare and as an `A) … (recommended)` label; 20 common phrasings, including the four plan-ceo-review mode names, map to their keys; negative controls (unregistered ids, a question with no options, a label naming no option, a mid-sentence cue, another question's key, a number that looks like a selector) stay unmapped. Two capture tests run the real bin: hook labels become keys with `choice_label` kept, while agent rows (including a `HOLD SCOPE` auto-decide row), ad-hoc ids, unmatched labels and free text are untouched, and a picked label follows a key recommendation. Both capture tests fail against the old bin.
+
+### Gate integrity
+- No test expectation, threshold, budget, timeout, trial count or case kind changed. Existing tests that pin label rows from the agent, auto-decide and Codex import paths pass unchanged.
+
+## [1.91.222.0] - 2026-10-25
+
+**Five flaky tests now fail only for real defects, and a judge reply that isn't valid JSON says so in its failure line (daily flaky-fix pass, GSTA-91).** `context-restore-fragment-match` failed 0/3 on main cb99df4, although every run restored the right checkpoint: to find the checkpoint whose title matches, the agent often prints all three, and the case failed whenever another checkpoint's body appeared anywhere in the transcript. `test/gstack-ff-update.test.ts` counted every `sleep 300` on the machine, so heartbeats from parallel test files changed its count (remote pregate at #3412's head: expected "1", received "3"). The `GSTACK_CAPY=1` install-admission case in `test/gstack-skill-start.test.ts` makes ten `gstack-skill-start` spawns under bun's 5 s default timeout and took 3.6–5.7 s on a 4-core machine. The QA-only fixture's late scenarios mocked a 1000 ms capture budget but ran each attempt's real setup inside it. The setup-browser-cookies workflow judge in train-4 gate 40173759210 returned JSON with an unescaped quote in its reasoning, and the flake record read only "JSON Parse error: Expected '}'".
+
+### Fixed
+- **`test/helpers/context-restore-fragment.ts`** (new): `fragmentRestoreVerdict` reads only the agent's report (top-level assistant text and the final result), never tool output. The report must carry the payments checkpoint's body and no other checkpoint's body. `test/skill-e2e-context-skills.test.ts` uses it for the fragment-match case.
+- **`test/helpers/llm-judge.ts`**: when a judge response is not valid JSON, `callJudge` throws a `SyntaxError` that names the judge output, its stop reason and length, and the retained `llm-judge-response-parse-error` event. It still fails closed and parses no score; the request is unchanged.
+
+### Tests
+- `test/context-restore-fragment-verdict.test.ts` replays the three captured false reds from main (`test/fixtures/context-restore-fragment-match-captures.json`) as passes. Controls that must fail: captures rewritten to resume alpha or omega, synthetic runs that read and resume only alpha or omega, resume summaries that mix two checkpoints, and a payments body that appears only in tool output. Restoring the old any-byte surface fails all four positive controls. Focused paid runs of the case: 0/3 on main, 3/3 after (about $0.11 a trial).
+- `test/llm-judge-frontier.test.ts` replays the exact captured cookie-judge reply: it fails closed with no score, one request and the new diagnosis. It fails on the previous helper.
+- `test/gstack-ff-update.test.ts`: the reap case runs `gstack-ff-update` with a heartbeat interval unique to the test process and counts only sleepers of that interval, starting from zero. The old count failed with a loop starting unrelated `sleep 300` processes beside it (expected "5", received "14"); the new one passes, and still fails when the bin stops reaping its sleeper.
+- `test/qa-only-fixture.test.ts`: in the late scenarios the fixture clock stands still from the start of each attempt until its runner starts, so the real setup work (directory copies and five `git` calls) costs no fixture time. The runner and cleanup stay on the real clock, as the deadline-evidence check requires. Each late capture must now get exactly the 1000 ms budget. A new `late-slow-retry` scenario stalls each attempt's setup for 1100 ms, longer than the whole budget, and its attempt 2 still has to succeed. Before, a loaded machine either used up the work budget in setup (`QA-only work budget exhausted`) or Bun's 1580 ms deadline killed a setup `git` child (exit `null` at `skill-e2e-qa-workflow.test.ts:133`); this is what failed late-retry in the 2026-10-24 and 2026-10-25 remote pregates. Under 24 busy loops on 4 cores: late-retry 1/20 → 20/20, late-failure-retry 2/20 → 20/20, late-timeout 6/20 → 20/20, late-slow-retry 20/20. Reverting the clock fails all four late scenarios unloaded; reverting the deadline fails `late-slow-retry`.
+- `test/gstack-skill-start.test.ts`: the `GSTACK_CAPY=1` install-admission case gets an explicit 20 s timeout. Measured on a 4-core Capy machine with a temporary 60 s cap: 3.6–5.7 s through Capy's git shim (9 of 10 over 5 s), 2.6–4.9 s with the real git binary. Four of its ten spawns are full admissions at about 800 ms each.
+
+### Gate integrity
+- One test budget changed, approved by GStack Minder on 2026-10-25: the install-admission case's timeout goes from bun's 5 s default to 20 s, about 3.5 times the 5.7 s worst case measured above. No assertion changed.
+- The QA-only late scenarios register each attempt with its real outer budget (1000 + 80 + 5000 = 6080 ms, the value `outerMs` already asserts) instead of a shortened 1580 ms Bun deadline, which left cleanup 500 ms of real time under load. Measured child totals with the fix under 24 busy loops, 20 runs each: late-timeout 2.4–2.7 s, late-retry 4.1–4.3 s, late-failure-retry 4.7–5.0 s, late-slow-retry 6.2–6.6 s, against the parent's unchanged 15 s spawn timeout. The mocked 1000 ms capture budget, the 80 ms drain and every existing assertion are unchanged. Verdict: justified, because the deadline was a fixture artifact shorter than the budget the case declares, and the clock freeze, not the deadline, is what removes the load sensitivity.
+- `context-restore-fragment-match`: the failure surface moves from the whole transcript to the agent's report. Alpha or omega bytes in tool output no longer fail the case; the payments body in tool output no longer counts as loaded, so the report must show it. A run that resumes another checkpoint, or mixes two, still fails. `test/test-of-test-ratchet.test.ts` gains one BASELINE entry, the replay test, with its reason, and the new helper joins the touchfiles of the nine context-skills cases, as the derived-closure rule requires.
+- `scripts/harness-version.json` records `test/helpers/llm-judge.ts` as non-behavioral at HARNESS_VERSION 6: only a parse failure's message changes, before any score is parsed, so no case's verdict or pass-rate series changes.
+- No threshold, trial count, case kind, rubric, judge request or prompt changed.
+
+## [1.91.221.0] - 2026-10-25
+
+**The session-start updater no longer skips because of its own session's workflow lease, and the autostash tests stop failing their first pass in Free Tests.** Since #3385, `gstack-session-update` skips the fast-forward while another session holds a workflow lease, and it means to ignore the session that started it (`--exclude-ancestry-of`). The check runs in the forked background updater, though, and passed the hook's own `$$`, which has exited by then, so the ancestry walk excluded nothing. A Claude Code process that had already run a skill (and so held a lease) and then started a session again in the same process (`/clear`, resume or compaction) logged `SKIP leased pid=<its own pid>` and never updated. `test/session-update-autostash.test.ts` hit the same gap plus a fixture leak: its first two fixtures spawned the updater with the whole parent environment, so it read the test preload's `GSTACK_LEASE_DIR`, which every test file in the same Bun process shares. A lease another file left there gave "Received: SKIP leased pid=3196/1m" where UPDATING or PULL_FAILED was expected, in 13 Free Tests runs since #3385 merged (each passed on the serial retry).
+
+### Fixed
+- **`bin/gstack-session-update`**: before it forks, the hook resolves its caller's harness pid with `gstack_harness_pid "$PPID"` (`bin/gstack-harness-pid.sh`), the same derivation `bin/gstack-skill-start` uses for the pid it leases, and the background lease check excludes that pid's ancestry. Leases held by other live sessions still skip the update. `bin/gstack-upgrade` runs its check in the foreground, so its `$$` is alive and it is unchanged.
+
+### Tests
+- `test/session-update-autostash.test.ts`: every fixture now runs the updater with its own `GSTACK_STATE_ROOT` (the top of the state-root chain, instead of `GSTACK_STATE_DIR`, which an inherited `GSTACK_HOME` outranks) and its own `GSTACK_LEASE_DIR`. A negative control holds a live lease in the shared preload lease directory, confirms the ambient environment sees it, and shows the fixture still updates; it fails with "SKIP leased" without the fixture change. A positive case shows the fixture's own live lease still skips the update.
+- `test/session-update-stages.test.ts`: the starting session's own lease (the test process, behind a `/bin/sh -c` wrapper as a hook runner would use) no longer blocks the update. It fails on main with `SKIP leased`. The other session-update test files already pass an explicit environment, so their leases resolve to the fixture install's own git dir.
+
+### Gate integrity
+- No threshold, budget, timeout, trial count, case kind or assertion changed. The lease rule is narrower in one respect only: the starting session's own harness and its ancestors no longer count as holders, which is what `--exclude-ancestry-of` promised.
+
 ## [1.91.220.0] - 2026-10-25
 
 **A design doc edited in host plan mode no longer goes stale for the plan reviews (#542).** When /office-hours ran inside Claude Code's plan mode, the design doc lived twice: gstack's copy under `~/.gstack/projects/<slug>/` and the host's plan file under `.claude/plans/`. Later edits went only to the plan file, so /plan-ceo-review and /plan-eng-review read the older gstack copy and called it the source of truth. The gstack design doc stays the single source (Garry, Oct 17: "designdoc"). Edits made in the plan copy are now written through to it, and every plan review is warned when the copy is newer.

NOTICE.md

@@ -1,9 +1,9 @@
 # Third-Party Notices
 
-gstack is MIT. The files below contain material derived from Apache-2.0 works,
-rewritten in gstack's voice and modified. Rewriting is not an exemption: the
-license text is in `licenses/Apache-2.0.txt`, and each derived file carries a
-notice that it was changed. Unmodified copies (the rule-registry fixture, the
+gstack is MIT. The files below contain material derived from Apache-2.0 and MIT
+works, rewritten in gstack's voice and modified. Rewriting is not an exemption:
+the license texts are in `licenses/`, and each file derived from an Apache-2.0
+work carries a notice that it was changed. Unmodified copies (the rule-registry fixture, the
 license text) carry no header and are listed here instead.
 
 ## impeccable — Copyright Paul Bakaus — Apache License 2.0
@@ -29,6 +29,27 @@ against the checksum pinned in `lib/design-detect-contract.ts`, and recorded in
 the egress ledger first. gstack does not audit the engine's network behavior; the
 wrapper refuses URL targets so gstack never asks it to touch the network.
 
+## ai-design-skills and redesign-skill — Copyright (c) 2026 Elaya — MIT License
+
+https://github.com/elayadesign/ai-design-skills (`skills/landing-page-design`,
+`skills/design-to-motion`, at 38bbe576) and
+https://github.com/elayadesign/redesign-skill (`skills/redesign-existing-projects`,
+at 64627f6e). License text: `licenses/MIT-elaya.txt`.
+
+Derived, modified (rules rewritten in gstack's voice and reconciled with its
+design rules; Elaya's fixed fonts, colors, spacing values, icon sets and
+signature components are not carried over):
+
+- `design-html/sections/craft.md.tmpl`: the Persuade intake, page archetypes,
+  page structure, conversion and copy rules, indexing and build order from
+  landing-page-design Part A; the content-realism, ship-floor, scale-snapping,
+  line-break and motion rules from landing-page-design Part B, the website
+  context of design-to-motion and the redesign-skill audit; and the generated
+  `design-html/SKILL.md` / `sections/craft.md`.
+- `lib/design-craft.ts` and `bin/gstack-design-craft`: the content-realism and
+  ship-floor tells, turned into a free check.
+- `design-shotgun/SKILL.md.tmpl`: page archetypes as a variant axis.
+
 ## DESIGN.md specification — Copyright Google LLC — Apache License 2.0
 
 https://github.com/google-labs-code/design.md

VERSION

@@ -1 +1 @@
-1.91.220.0
+1.91.226.0

agents-digest/gstack-AGENTS.md

@@ -1,4 +1,4 @@
-# gstack digest v1.91.220.0 — regenerate/re-copy after upgrading gstack
+# gstack digest v1.91.226.0 — regenerate/re-copy after upgrading gstack
 
 Behavioral rules from gstack (https://github.com/garrytan/gstack), compressed
 for agent hosts without a full skill install. The full skills add workflows,

bin/gstack-design-craft

@@ -0,0 +1,37 @@
+#!/usr/bin/env -S bun --no-env-file --config=/dev/null
+// gstack-design-craft — free craft check for a generated page (/design-html Step 4).
+//
+// Usage: gstack-design-craft <file.html>
+//
+// Prints one "CRAFT_CHECK: <id> <count> <fix>" line per tell, then
+// "CRAFT_RESULT: clean" (exit 0) or "CRAFT_RESULT: <n> issue(s)" (exit 1).
+// Exit 2: usage error, a non-HTML file, an unreadable or empty file. Head and
+// navigation rules run only on a full HTML document; a fragment gets the copy
+// and code rules alone.
+
+import { readFileSync } from 'fs';
+import { checkCraft, formatCraft } from '../lib/design-craft';
+
+const file = process.argv[2];
+if (!file) {
+  console.error('Usage: gstack-design-craft <file.html>');
+  process.exit(2);
+}
+if (!/\.html?$/i.test(file)) {
+  console.error(`gstack-design-craft: ${file} is not an HTML file; check the rendered HTML preview instead`);
+  process.exit(2);
+}
+let source: string;
+try {
+  source = readFileSync(file, 'utf-8');
+} catch (err) {
+  console.error(`gstack-design-craft: cannot read ${file}: ${(err as Error).message}`);
+  process.exit(2);
+}
+if (source.trim() === '') {
+  console.error(`gstack-design-craft: ${file} is empty; nothing was checked`);
+  process.exit(2);
+}
+const findings = checkCraft(source);
+console.log(formatCraft(findings));
+process.exit(findings.length === 0 ? 0 : 1);

bin/gstack-question-log

@@ -18,8 +18,13 @@
 #   door_type          — one-way | two-way
 #                        (optional — looked up from registry if omitted)
 #   options_count      — number of options presented (positive integer)
-#   user_choice        — key user selected (free string; registry-options preferred)
-#   recommended        — option key the agent recommended (optional)
+#   user_choice        — key user selected (free string; registry-options preferred).
+#                        On hook-captured events (source=hook) for a registered
+#                        question_id, the clicked UI label is mapped to its
+#                        registered option key (#2623) and kept in choice_label.
+#                        Unmatched labels and other sources stay as given.
+#   recommended        — option key the agent recommended (optional; mapped
+#                        the same way on hook-captured events)
 #   followed_recommendation — bool (optional — computed if both present)
 #   session_id         — stable session identifier
 #   ts                 — ISO 8601 timestamp (auto-injected if missing)
@@ -46,8 +51,9 @@ INPUT="$1"
 TMPERR=$(mktemp "${TMPDIR:-/tmp}/gstack-question-log.XXXXXX") || { echo "gstack-question-log: could not create a temp file in ${TMPDIR:-/tmp}, so the question was not logged. Fix: point TMPDIR at a writable directory." >&2; exit 1; }
 trap 'rm -f "$TMPERR"' EXIT
 set +e
-VALIDATED=$(printf '%s' "$INPUT" | GSTACK_LIB_DIR="$SCRIPT_DIR/../lib" bun --no-env-file --config=/dev/null -e "
+VALIDATED=$(printf '%s' "$INPUT" | GSTACK_LIB_DIR="$SCRIPT_DIR/../lib" GSTACK_SCRIPTS_DIR="$SCRIPT_DIR/../scripts" bun --no-env-file --config=/dev/null -e "
 const { hasInjection } = await import(process.env.GSTACK_LIB_DIR + '/jsonl-store.ts');
+const { resolveOptionKey } = await import(process.env.GSTACK_SCRIPTS_DIR + '/psychographic-signals.ts');
 const path = require('path');
 const raw = await Bun.stdin.text();
 let j;
@@ -173,6 +179,23 @@ if (j.recommended !== undefined) {
   if (j.recommended.length > 64) j.recommended = j.recommended.slice(0, 64);
 }
 
+// Map the clicked UI label to its registered option key (#2623). SIGNAL_MAP
+// and question preferences are keyed by the registry's option keys, but the
+// PostToolUse hook only sees the label. Hook events and registered ids only;
+// an unmatched or ambiguous label stays as given. Agent-written rows keep
+// what the agent logged: skills such as plan-ceo-review log full mode names
+// that their own readers match on.
+if (j.source === 'hook') {
+  const key = resolveOptionKey(j.question_id, j.user_choice);
+  if (key !== undefined && key !== j.user_choice) {
+    j.choice_label = j.user_choice;
+    j.user_choice = key;
+  }
+  if (j.recommended !== undefined) {
+    j.recommended = resolveOptionKey(j.question_id, j.recommended) ?? j.recommended;
+  }
+}
+
 // followed_recommendation — compute if both sides present. An __unknown__
 // choice means extraction failed, not that the user rejected the
 // recommendation — leave the field absent so metrics can't be poisoned.

bin/gstack-session-update

@@ -252,6 +252,13 @@ if [ -f "$THROTTLE_FILE" ]; then
   fi
 fi
 
+# The caller's own session never counts as a lease holder (#3363). Resolve its
+# harness pid now, the same way bin/gstack-skill-start names the pid it leases:
+# once this hook exits, the background updater's $$ names a dead process and
+# an ancestry walk from it excludes nothing.
+. "$(cd "$(dirname "$0")" && pwd)/gstack-harness-pid.sh" 2>/dev/null || gstack_harness_pid() { printf '%s\n' "$1"; }
+CALLER_PID=$(gstack_harness_pid "$PPID")
+
 # ── Fork to background: zero latency on session start ──
 (
   # Prevent git from prompting for credentials (would hang the background process)
@@ -368,7 +375,7 @@ fi
   # retry runs under it, the rule bin/gstack-upgrade enforces as UPGRADE_BUSY.
   # The throttle file is not written, so the next session start tries again.
   if [ "${GSTACK_UPGRADE_FORCE:-}" != 1 ] && [ -x "$LEASE" ] \
-    && BUSY=$("$LEASE" live --root "$GSTACK_DIR" --exclude-ancestry-of $$ 2>/dev/null); then
+    && BUSY=$("$LEASE" live --root "$GSTACK_DIR" --exclude-ancestry-of "$CALLER_PID" 2>/dev/null); then
     log_entry "SKIP leased $(printf '%s\n' "$BUSY" | awk 'NF { printf "%spid=%s/%s", n++ ? " " : "", $1, $2 }')"
     exit 0
   fi

bin/gstack-slug

@@ -8,7 +8,7 @@
 #                                  Code sessions refuse eval/source)
 # Modes (not on the hot path; see `gstack-slug --help`):
 #   --identity                 also print PROJECT_REMOTE, PROJECT_ROOT, LEGACY_SLUG
-#   --stamp-checkpoint <file>  write remote:/project_root: into a checkpoint's frontmatter
+#   --stamp-checkpoint <file>  write remote:/project_root:/head: into a checkpoint's frontmatter
 #   --classify-checkpoints     read checkpoint paths on stdin, print "<status>\t<path>"
 #   --checkpoints-below        list other buckets' checkpoints saved below the cwd, plus pointers
 #   --point-from <dir> <file>  leave a pointer to <file> in <dir>'s bucket (nested-repo saves)
@@ -138,7 +138,7 @@ gstack-slug — print SLUG=<project slug> and BRANCH=<branch> for the current di
                                         field); no eval needed
   gstack-slug --identity                also print PROJECT_REMOTE (canonical, credential-free),
                                         PROJECT_ROOT and LEGACY_SLUG (pre-#3003 bucket, if different)
-  gstack-slug --stamp-checkpoint FILE   write remote:/project_root: into FILE's frontmatter
+  gstack-slug --stamp-checkpoint FILE   write remote:/project_root:/head: into FILE's frontmatter
   gstack-slug --classify-checkpoints    read checkpoint paths on stdin; print "<status><TAB><path>",
                                         status = match | match-root | foreign | unstamped
   gstack-slug --checkpoints-below       outside git: other buckets' checkpoints whose project_root
@@ -682,15 +682,18 @@ case "$MODE" in
     STAMP_TMP=$(mktemp "${MODE_ARG}.stamp-XXXXXX") || exit 1
     STAMP_WORKTREE=$(_git_toplevel "$PROJECT_DIR")
     [[ "$STAMP_WORKTREE" != "$ID_ROOT" ]] || STAMP_WORKTREE=""
-    if GRI_REMOTE="$ID_REMOTE" GRI_ROOT="$ID_ROOT" GRI_WORKTREE="$STAMP_WORKTREE" awk '
+    # The commit the save describes, so /context-restore can tell when work landed after it.
+    STAMP_HEAD=$(git -C "$PROJECT_DIR" rev-parse --verify -q HEAD 2>/dev/null) || STAMP_HEAD=""
+    if GRI_REMOTE="$ID_REMOTE" GRI_ROOT="$ID_ROOT" GRI_WORKTREE="$STAMP_WORKTREE" GRI_HEAD="$STAMP_HEAD" awk '
       function ids() {
         if (ENVIRON["GRI_REMOTE"] != "") print "remote: " ENVIRON["GRI_REMOTE"]
         print "project_root: " ENVIRON["GRI_ROOT"]
         if (ENVIRON["GRI_WORKTREE"] != "") print "worktree: " ENVIRON["GRI_WORKTREE"]
+        if (ENVIRON["GRI_HEAD"] != "") print "head: " ENVIRON["GRI_HEAD"]
       }
       NR == 1 { if ($0 == "---") { infm = 1; print; next } print "---"; ids(); print "---" }
       infm && $0 == "---" { ids(); print; infm = 0; next }
-      infm && (/^remote:/ || /^project_root:/ || /^worktree:/) { next }
+      infm && (/^remote:/ || /^project_root:/ || /^worktree:/ || /^head:/) { next }
       { print }
       END { if (NR == 0) { print "---"; ids(); print "---" } }' "$MODE_ARG" > "$STAMP_TMP"; then
       mv "$STAMP_TMP" "$MODE_ARG"

context-restore/SKILL.md

@@ -527,8 +527,50 @@ checkpoint(s) from another project; only this project's are listed." If it print
 `LEGACY_BUCKET`, relay that line. If the chosen file is `ROOT_DIFFERS`, add
 "Saved from another checkout of this repository at `{project_root}`." as info.
 
+**Check whether the branch moved since the save.** The stamp records the commit
+the save was made at in `head:`. Before presenting, compare it with this
+checkout. FILE is the checkpoint path chosen in Step 2, one of the paths Step 1
+printed:
+
+```bash
+FILE="<the FILE path printed above>"
+SAVED_HEAD=$(tr -d '\r' < "$FILE" 2>/dev/null | sed -n '1{/^---$/!q;d;}; /^---$/q; /^head:/{s/^head:[[:space:]]*//p;q;}')
+REPO_DIR=.
+git rev-parse --git-dir >/dev/null 2>&1 || REPO_DIR=$(tr -d '\r' < "$FILE" 2>/dev/null | sed -n '1{/^---$/!q;d;}; /^---$/q; /^worktree:/{s/^worktree:[[:space:]]*//p;q;}')
+[ -n "$REPO_DIR" ] || REPO_DIR=$(tr -d '\r' < "$FILE" 2>/dev/null | sed -n '1{/^---$/!q;d;}; /^---$/q; /^project_root:/{s/^project_root:[[:space:]]*//p;q;}')
+CURRENT_HEAD=$(git -C "${REPO_DIR:-.}" rev-parse --verify -q HEAD 2>/dev/null)
+if [ -z "$SAVED_HEAD" ]; then
+  echo "HEAD_UNTRACKED"
+elif [ -z "$CURRENT_HEAD" ]; then
+  echo "HEAD_NOT_CHECKED"
+elif [ "$SAVED_HEAD" = "$CURRENT_HEAD" ]; then
+  echo "HEAD_UNCHANGED"
+elif ! printf '%s\n' "$SAVED_HEAD" | grep -Eqx '[0-9a-f]{40}([0-9a-f]{24})?' \
+  || ! git -C "$REPO_DIR" cat-file -e "$SAVED_HEAD^{commit}" 2>/dev/null; then
+  echo "HEAD_UNKNOWN $SAVED_HEAD"
+elif git -C "$REPO_DIR" merge-base --is-ancestor "$SAVED_HEAD" "$CURRENT_HEAD"; then
+  echo "HEAD_MOVED $(git -C "$REPO_DIR" rev-list --count "$SAVED_HEAD..$CURRENT_HEAD")"
+  git -C "$REPO_DIR" log --oneline -10 "$SAVED_HEAD..$CURRENT_HEAD"
+else
+  echo "HEAD_DIVERGED $SAVED_HEAD"
+fi
+```
+
+- `HEAD_UNCHANGED` or `HEAD_NOT_CHECKED`: nothing to add.
+- `HEAD_MOVED <n>`: print `<n> commit(s) landed since this save:` and the commit
+  lines above the Remaining Work groups, and list **every** Remaining Work item
+  under **Verify first**: the saving session could not know about those commits,
+  so any step may already be done.
+- `HEAD_DIVERGED` or `HEAD_UNKNOWN`: print `This checkpoint was saved at {first 7
+  characters of head}, which is not in this checkout's history (saved on another
+  branch, rebased, reset, or not fetched).` and list every item under Verify first.
+- `HEAD_UNTRACKED`: print `This checkpoint predates commit tracking; restore cannot tell whether work landed after it.`
+  and sort as usual.
+
 **Sort Remaining Work by provenance.** Keep every item's original text and saved
-order, and drop nothing. Put an item under **Verify first** when it:
+order, and drop nothing. When the head check printed `HEAD_MOVED`, `HEAD_DIVERGED`
+or `HEAD_UNKNOWN`, skip the rules below: every item goes under Verify first,
+whatever its marker. Put an item under **Verify first** when it:
 - ends in `(path assumed)` or `(code read)`;
 - ends in `(path run)` but its text reports a failure;
 - ends in a parenthesised trailing marker that is **not** one of the five English
@@ -563,6 +605,7 @@ Status:      {status}
 {summary from saved file}
 
 ### Remaining Work
+{commit-tracking lines, if any}
 {legacy banner line, if it applies}
 
 Next steps
@@ -616,6 +659,8 @@ in the old bucket until the user copies them with `gstack-slug --adopt-legacy`.
 - **Never present another project's checkpoint as "latest".** Identity is the
   canonical remote (the root only when there is no remote); a mismatch is
   reported before any summary.
+- **A checkpoint describes the commit it was saved at.** When commits landed
+  after the save, every Remaining Work item is verified before it is run.
 - **"Most recent" means the filename `YYYYMMDD-HHMMSS` prefix**, not
   `ls -1t` (filesystem mtime). Filenames are stable across file-system
   operations; mtime is not.

context-restore/SKILL.md.tmpl

@@ -203,8 +203,50 @@ checkpoint(s) from another project; only this project's are listed." If it print
 `LEGACY_BUCKET`, relay that line. If the chosen file is `ROOT_DIFFERS`, add
 "Saved from another checkout of this repository at `{project_root}`." as info.
 
+**Check whether the branch moved since the save.** The stamp records the commit
+the save was made at in `head:`. Before presenting, compare it with this
+checkout. FILE is the checkpoint path chosen in Step 2, one of the paths Step 1
+printed:
+
+```bash
+FILE="<the FILE path printed above>"
+SAVED_HEAD=$(tr -d '\r' < "$FILE" 2>/dev/null | sed -n '1{/^---$/!q;d;}; /^---$/q; /^head:/{s/^head:[[:space:]]*//p;q;}')
+REPO_DIR=.
+git rev-parse --git-dir >/dev/null 2>&1 || REPO_DIR=$(tr -d '\r' < "$FILE" 2>/dev/null | sed -n '1{/^---$/!q;d;}; /^---$/q; /^worktree:/{s/^worktree:[[:space:]]*//p;q;}')
+[ -n "$REPO_DIR" ] || REPO_DIR=$(tr -d '\r' < "$FILE" 2>/dev/null | sed -n '1{/^---$/!q;d;}; /^---$/q; /^project_root:/{s/^project_root:[[:space:]]*//p;q;}')
+CURRENT_HEAD=$(git -C "${REPO_DIR:-.}" rev-parse --verify -q HEAD 2>/dev/null)
+if [ -z "$SAVED_HEAD" ]; then
+  echo "HEAD_UNTRACKED"
+elif [ -z "$CURRENT_HEAD" ]; then
+  echo "HEAD_NOT_CHECKED"
+elif [ "$SAVED_HEAD" = "$CURRENT_HEAD" ]; then
+  echo "HEAD_UNCHANGED"
+elif ! printf '%s\n' "$SAVED_HEAD" | grep -Eqx '[0-9a-f]{40}([0-9a-f]{24})?' \
+  || ! git -C "$REPO_DIR" cat-file -e "$SAVED_HEAD^{commit}" 2>/dev/null; then
+  echo "HEAD_UNKNOWN $SAVED_HEAD"
+elif git -C "$REPO_DIR" merge-base --is-ancestor "$SAVED_HEAD" "$CURRENT_HEAD"; then
+  echo "HEAD_MOVED $(git -C "$REPO_DIR" rev-list --count "$SAVED_HEAD..$CURRENT_HEAD")"
+  git -C "$REPO_DIR" log --oneline -10 "$SAVED_HEAD..$CURRENT_HEAD"
+else
+  echo "HEAD_DIVERGED $SAVED_HEAD"
+fi
+```
+
+- `HEAD_UNCHANGED` or `HEAD_NOT_CHECKED`: nothing to add.
+- `HEAD_MOVED <n>`: print `<n> commit(s) landed since this save:` and the commit
+  lines above the Remaining Work groups, and list **every** Remaining Work item
+  under **Verify first**: the saving session could not know about those commits,
+  so any step may already be done.
+- `HEAD_DIVERGED` or `HEAD_UNKNOWN`: print `This checkpoint was saved at {first 7
+  characters of head}, which is not in this checkout's history (saved on another
+  branch, rebased, reset, or not fetched).` and list every item under Verify first.
+- `HEAD_UNTRACKED`: print `This checkpoint predates commit tracking; restore cannot tell whether work landed after it.`
+  and sort as usual.
+
 **Sort Remaining Work by provenance.** Keep every item's original text and saved
-order, and drop nothing. Put an item under **Verify first** when it:
+order, and drop nothing. When the head check printed `HEAD_MOVED`, `HEAD_DIVERGED`
+or `HEAD_UNKNOWN`, skip the rules below: every item goes under Verify first,
+whatever its marker. Put an item under **Verify first** when it:
 - ends in `(path assumed)` or `(code read)`;
 - ends in `(path run)` but its text reports a failure;
 - ends in a parenthesised trailing marker that is **not** one of the five English
@@ -239,6 +281,7 @@ Status:      {status}
 {summary from saved file}
 
 ### Remaining Work
+{commit-tracking lines, if any}
 {legacy banner line, if it applies}
 
 Next steps
@@ -292,6 +335,8 @@ in the old bucket until the user copies them with `gstack-slug --adopt-legacy`.
 - **Never present another project's checkpoint as "latest".** Identity is the
   canonical remote (the root only when there is no remote); a mismatch is
   reported before any summary.
+- **A checkpoint describes the commit it was saved at.** When commits landed
+  after the save, every Remaining Work item is verified before it is run.
 - **"Most recent" means the filename `YYYYMMDD-HHMMSS` prefix**, not
   `ls -1t` (filesystem mtime). Filenames are stable across file-system
   operations; mtime is not.

context-save/SKILL.md

@@ -542,12 +542,13 @@ files_modified:
 
 The `files_modified` list comes from `git status --short` (both staged and unstaged
 modified files). Use relative paths from the repo root. Do not write `remote:`,
-`project_root:` or `worktree:` yourself; the stamp step below adds them.
+`project_root:`, `worktree:` or `head:` yourself; the stamp step below adds them.
 
 Then stamp the project identity. Several projects can share one checkpoints
 directory, so `/context-restore` needs to know which project wrote each file. The helper adds `remote:` (canonical origin:
-lowercased host, no credentials, port or `.git`), `project_root:` and, in a
-linked worktree, `worktree:` to the frontmatter. When the session started in
+lowercased host, no credentials, port or `.git`), `project_root:`, in a linked
+worktree `worktree:`, and `head:` (the commit this save describes, so
+`/context-restore` can tell when work landed after it) to the frontmatter. When the session started in
 another directory that files elsewhere (a non-git parent of this repo), it also
 leaves a pointer there so `/context-restore` in that directory finds this save:
 

context-save/SKILL.md.tmpl

@@ -217,12 +217,13 @@ files_modified:
 
 The `files_modified` list comes from `git status --short` (both staged and unstaged
 modified files). Use relative paths from the repo root. Do not write `remote:`,
-`project_root:` or `worktree:` yourself; the stamp step below adds them.
+`project_root:`, `worktree:` or `head:` yourself; the stamp step below adds them.
 
 Then stamp the project identity. Several projects can share one checkpoints
 directory, so `/context-restore` needs to know which project wrote each file. The helper adds `remote:` (canonical origin:
-lowercased host, no credentials, port or `.git`), `project_root:` and, in a
-linked worktree, `worktree:` to the frontmatter. When the session started in
+lowercased host, no credentials, port or `.git`), `project_root:`, in a linked
+worktree `worktree:`, and `head:` (the commit this save describes, so
+`/context-restore` can tell when work landed after it) to the frontmatter. When the session started in
 another directory that files elsewhere (a non-git parent of this repo), it also
 leaves a pointer there so `/context-restore` in that directory finds this save:
 

design-html/SKILL.md

@@ -366,6 +366,7 @@ sections. Read a section in full before doing its step; do not work from memory.
 |------|-------------------|
 | analyzing the design or making any layout/visual decision (Step 1 onward) — the UX-principles doctrine governs every design choice | `sections/doctrine.md` |
 | writing the finalized HTML in Step 3 — the Pretext wiring patterns and API cheatsheet are the required reference for all text-layout code | `sections/pretext-patterns.md` |
+| writing the Step 1 implementation spec — the Persuade intake and archetype, and the craft floor every page ships with | `sections/craft.md` |
 | the Setup probe printed DESIGN_DETECTOR_INSTALL_OFFER — ask the user once whether gstack may download impeccable's engine (checksum-pinned, receipted) before any other step | `sections/detector-install-offer.md` |
 
 ---
@@ -535,6 +536,9 @@ After routing, output a brief context summary:
 
 ## Step 1: Design Analysis
 
+> **STOP.** Before writing the Step 1 implementation spec — the Persuade intake and archetype, and the craft floor every page ships with, Read `~/.claude/skills/gstack/design-html/sections/craft.md` and execute it
+> in full. Do not work from memory — that section is the source of truth for this step.
+
 1. If `$D` is available (`DESIGN_READY`), extract a structured implementation spec:
 ```bash
 $D prompt --image <approved-variant.png> --output json
@@ -658,6 +662,7 @@ For framework output, save to:
 - `prefers-color-scheme` media query for dark mode
 - `prefers-reduced-motion` for animation respect
 - Real content extracted from the mockup (never lorem ipsum)
+- The craft section's ship floor, scale-snapped values and motion rules
 
 **Never include by default (AI slop blacklist):** an approved mockup that carries one, a DESIGN.md blessing, or an explicit user ask overrides it; say the tradeoff once.
 - Purple/blue gradients as default <!-- ai-color-palette -->
@@ -726,6 +731,8 @@ _DJ=$(mktemp "${TMPDIR:-/tmp}/gstack-detect.XXXXXX"); bun --no-env-file --config
 
 Exit 2 → one surgical fix pass over the non-advisory rules in the `DETECT_TOP` block, then scan once more. Whatever remains, present the page with those findings listed as accepted-with-reason: a pattern the approved mockup contains, a value DESIGN.md's tokens bless or a pattern its Decisions Log or Do's and Don'ts records as intentional, or an inline `<!-- impeccable-disable <rule>: <reason> -->` the user agreed to. One pass, not a loop. Any other first line from the probe: skip, no ceremony.
 
+Then, always: the craft section's check, `~/.claude/skills/gstack/bin/gstack-design-craft <path-to-finalized.html>`, one fix pass and one rerun, as above.
+
 ### Verification Screenshots
 
 Take verification screenshots at 3 viewports. One `gstack-render` call serves

design-html/SKILL.md.tmpl

@@ -156,6 +156,8 @@ After routing, output a brief context summary:
 
 ## Step 1: Design Analysis
 
+{{SECTION:craft}}
+
 1. If `$D` is available (`DESIGN_READY`), extract a structured implementation spec:
 ```bash
 $D prompt --image <approved-variant.png> --output json
@@ -278,6 +280,7 @@ For framework output, save to:
 - `prefers-color-scheme` media query for dark mode
 - `prefers-reduced-motion` for animation respect
 - Real content extracted from the mockup (never lorem ipsum)
+- The craft section's ship floor, scale-snapped values and motion rules
 
 **Never include by default (AI slop blacklist):** an approved mockup that carries one, a DESIGN.md blessing, or an explicit user ask overrides it; say the tradeoff once.
 - Purple/blue gradients as default <!-- ai-color-palette -->
@@ -338,6 +341,8 @@ kill <PID> 2>/dev/null || true
 
 {{DESIGN_DETECTOR:gate}}
 
+Then, always: the craft section's check, `{{BIN_DIR}}/gstack-design-craft <path-to-finalized.html>`, one fix pass and one rerun, as above.
+
 ### Verification Screenshots
 
 Take verification screenshots at 3 viewports. One `gstack-render` call serves

design-html/sections/craft.md

@@ -0,0 +1,72 @@
+<!-- AUTO-GENERATED from craft.md.tmpl — do not edit directly -->
+<!-- Regenerate: bun run gen:skill-docs -->
+## Craft: Persuade pages and the craft floor
+
+**Persuade pages** applies when the page exists to get one action from a visitor: a landing, marketing, pricing, waitlist, campaign or comparison page. Write `Surface: Persuade` or `Surface: other` as the spec's first line; on `other`, skip to Craft floor, which applies to every page. An approved mockup wins on structure and visuals; this section fills only what it leaves open.
+
+### Persuade pages
+
+**One offer → one audience → one primary action.**
+
+**Intake.** When the context does not settle them, ask one AskUserQuestion with at most three multiple-choice questions: the primary action, the traffic source, and published or design reference. Assume the rest in one `Assumed:` line the user corrects in the refinement loop:
+- the conversion and the offer: exactly what they get
+- who it is for, the problem they came with, their top three objections, what they already know
+- the proof the user can supply (named customers, attributed quotes, real numbers) and the risk terms the business offers
+
+**Archetype.** Pick one and say why in the spec's layout type: **classic hero + sections** (default; the product reads from one screenshot), **long-form story** (the visitor needs educating or is skeptical), **minimal conversion** (high-intent traffic or a small offer), **comparison** (search intent is "X vs Y").
+
+**Structure.** Above the fold at 375×667 and on desktop, inside the hero budget: a headline naming outcome and audience, one sentence on how, one primary CTA, one proof signal only when real proof exists (named logos or one attributed quote; a metric row stays banned), one product visual. Below it, choose modules and order from the archetype and the objections. The long-form menu: problem → solution, three to five benefits, how it works in three steps, proof, an FAQ of six to twelve real objections, the risk reversal, a final CTA identical to the top. A minimal page stays minimal; a comparison page leads with the comparison.
+
+**Conversion rules.**
+- A page behind an ad repeats the ad's promise and tone in the hero.
+- One primary CTA; nothing competes with it above the fold.
+- Benefit first and specific: "Cut weekly reporting from 4 hours to 15 minutes", never "save time".
+- Show the risk reversal the business offers (free plan, trial, no card, cancel anytime). Never invent one.
+- Proof sits beside the claim it supports. A high-friction offer moves the FAQ up.
+
+**The primary action works.** A real destination (supplied URL or form endpoint) and every state: invalid (inline error, input kept), pending (control disabled so it cannot submit twice, announced through `aria-live`), failure (what happened, input kept, retry) and success (what happens next). With no backend, simulate the states and say so in a code comment and your summary.
+
+**Copy.** A CTA is a verb plus what they get ("Start free trial"). Draft two or three headlines in the product's voice and keep the most specific one that names outcome and audience; "{outcome} without {pain}" is a test, not a template.
+
+**Indexing.** Record the publication decision; keep the site's policy when known. A page the user keeps out of search gets `<meta name="robots" content="noindex">`; an indexed page gets a real title and description, and its FAQ ships as FAQPage JSON-LD.
+
+**Build order.** One section at a time, top down; a refinement edits one section.
+
+### Craft floor (every page)
+
+**Claims come from the user, never from you.** Testimonials, customer names and logos, results, uptime, prices and guarantees appear only when supplied. Missing proof: leave the slot out, or mark it `[Needs proof: …]` in the warning color so it cannot ship by accident. A believable invented claim is worse than an obvious placeholder.
+
+**Illustrative content looks real.** Sample data in product shots and UI (table rows, chart values, people in a demo inbox):
+- No placeholder people or brands (John Doe, Jane Doe, Acme, Nexus, SmartFlow); distinct avatars; varied dates.
+- No round sample numbers (99.99%, 50%, 1,000+ customers, 10x). In a claim, a round number signals it was made up: use the supplied figure or drop it.
+- No buzzwords: elevate, seamless, effortless, supercharge, streamline, revolutionize, unlock, empower, unleash, next-gen, game-changer, delve, tapestry, "in the world of".
+- Sentence-case headings. Errors say what happened and what to do, never "Oops!". No exclamation marks in success messages.
+
+**Ship floor.** Head tags apply to a full HTML document, not a component or fragment. Every full document has `<title>`, a meta description, a favicon (an inline SVG data URI from the brand mark when no file is supplied), and on a page meant for sharing `og:title` and `og:description`; `og:image` only from a supplied URL. A "Skip to content" link is the first focusable element, hidden until focused, targeting the `id` of `<main>`. Landmarks (`<header>`, `<nav>`, `<main>`, `<footer>`); alt text on meaningful images, `alt=""` on decorative ones. A marketing footer links privacy and terms at supplied URLs. Required fields and email format validate inline, wired with `aria-invalid` and `aria-describedby`. Every link goes somewhere real: no `href="#"`, no made-up path; a control with no target renders disabled. Navigation to other pages marks the current one `aria-current="page"`; in-page section links do not.
+
+**Values come from the scale.** Every font size, space and radius is a step of the system's scale (DESIGN.md, else the spec), chosen by role. An off-scale value snaps to the nearest step that keeps its role, usually the one below, with that step's line height; never below 16px body text or a 44px touch target, and an approved mockup's measured values stand. A shape nested in another with a gap under 32px gets inner radius = outer − gap when that exceeds 2px.
+
+**Line breaks.** Browser-wrapped headings get `text-wrap: balance`, body `text-wrap: pretty`. Break display lines where the thought breaks; never one word alone on the last line.
+
+**Motion.** These govern the one authored motion moment and state transitions; they license no more motion.
+- Content is visible by default; script adds the hidden start state only after the observer is set up (`.motion-ready` on `<html>`), so a script failure or a crawler sees everything and the CTA never waits on an animation.
+- Reveal with `IntersectionObserver` (`threshold: 0`, bottom `rootMargin` about −20%, once). Never an unthrottled `window.addEventListener('scroll')`; scroll-linked effects read scroll inside `requestAnimationFrame`.
+- Animate `transform`, `opacity`, `filter`, `clip-path` only; blur 14px or less.
+- No object grows from `scale(0)` (start at 0.85 or more); only lines and bars grow from zero. A data bar fills as a left-to-right `clip-path` wipe, 0.7–1.1s.
+- Stagger in reading order inside one section (30–80ms), never across the page.
+- Curves are named tokens: DESIGN.md's, else `cubic-bezier(0.16, 1, 0.3, 1)` for entrances and `cubic-bezier(0.65, 0, 0.35, 1)` for wipes; `linear` only for constant motion.
+- Loops pause off-screen. `prefers-reduced-motion: reduce` shows the final state. Full-height sections use `min-height: 100dvh`, not `100vh`.
+
+### Craft check
+
+At Step 4, after the Slop Gate and before the screenshots, run the free checker on a vanilla HTML output (for framework output, say it is unavailable):
+
+```bash
+~/.claude/skills/gstack/bin/gstack-design-craft <path-to-finalized.html>
+```
+
+It prints one `CRAFT_CHECK: <id> <count> <fix>` line per tell (the realism rules above, `[Needs proof:` markers, scroll listeners, `100vh`, `transition: all`, `href="#"`, `alert()`, missing alt; on a full document a missing title, description, favicon, working skip link, or `aria-current` on route navigation), then `CRAFT_RESULT: clean` or `CRAFT_RESULT: <n> issue(s)`. No `CRAFT_RESULT` line: say the check was unavailable and continue. Same contract as the Slop Gate: one surgical fix pass, one rerun, the rest listed as accepted-with-reason (a real customer named Nexus, a supplied 50% figure). Rerun once when the refinement loop exits; report only new lines.
+
+**Before this can ship.** Open the Step 4 hand-off, and repeat before Step 5's "Copy to project": every `[Needs proof: …]` marker, the dependencies the page cannot own (legal URLs, form endpoint, og:image, indexing, analytics and cookie consent where required), the accepted exceptions, and every claim the check made you change.
+
+Adapted from Elaya's [ai-design-skills](https://github.com/elayadesign/ai-design-skills) and [redesign-skill](https://github.com/elayadesign/redesign-skill) (MIT, © 2026 Elaya), rewritten for gstack. See NOTICE.md.

design-html/sections/craft.md.tmpl

@@ -0,0 +1,70 @@
+## Craft: Persuade pages and the craft floor
+
+**Persuade pages** applies when the page exists to get one action from a visitor: a landing, marketing, pricing, waitlist, campaign or comparison page. Write `Surface: Persuade` or `Surface: other` as the spec's first line; on `other`, skip to Craft floor, which applies to every page. An approved mockup wins on structure and visuals; this section fills only what it leaves open.
+
+### Persuade pages
+
+**One offer → one audience → one primary action.**
+
+**Intake.** When the context does not settle them, ask one AskUserQuestion with at most three multiple-choice questions: the primary action, the traffic source, and published or design reference. Assume the rest in one `Assumed:` line the user corrects in the refinement loop:
+- the conversion and the offer: exactly what they get
+- who it is for, the problem they came with, their top three objections, what they already know
+- the proof the user can supply (named customers, attributed quotes, real numbers) and the risk terms the business offers
+
+**Archetype.** Pick one and say why in the spec's layout type: **classic hero + sections** (default; the product reads from one screenshot), **long-form story** (the visitor needs educating or is skeptical), **minimal conversion** (high-intent traffic or a small offer), **comparison** (search intent is "X vs Y").
+
+**Structure.** Above the fold at 375×667 and on desktop, inside the hero budget: a headline naming outcome and audience, one sentence on how, one primary CTA, one proof signal only when real proof exists (named logos or one attributed quote; a metric row stays banned), one product visual. Below it, choose modules and order from the archetype and the objections. The long-form menu: problem → solution, three to five benefits, how it works in three steps, proof, an FAQ of six to twelve real objections, the risk reversal, a final CTA identical to the top. A minimal page stays minimal; a comparison page leads with the comparison.
+
+**Conversion rules.**
+- A page behind an ad repeats the ad's promise and tone in the hero.
+- One primary CTA; nothing competes with it above the fold.
+- Benefit first and specific: "Cut weekly reporting from 4 hours to 15 minutes", never "save time".
+- Show the risk reversal the business offers (free plan, trial, no card, cancel anytime). Never invent one.
+- Proof sits beside the claim it supports. A high-friction offer moves the FAQ up.
+
+**The primary action works.** A real destination (supplied URL or form endpoint) and every state: invalid (inline error, input kept), pending (control disabled so it cannot submit twice, announced through `aria-live`), failure (what happened, input kept, retry) and success (what happens next). With no backend, simulate the states and say so in a code comment and your summary.
+
+**Copy.** A CTA is a verb plus what they get ("Start free trial"). Draft two or three headlines in the product's voice and keep the most specific one that names outcome and audience; "{outcome} without {pain}" is a test, not a template.
+
+**Indexing.** Record the publication decision; keep the site's policy when known. A page the user keeps out of search gets `<meta name="robots" content="noindex">`; an indexed page gets a real title and description, and its FAQ ships as FAQPage JSON-LD.
+
+**Build order.** One section at a time, top down; a refinement edits one section.
+
+### Craft floor (every page)
+
+**Claims come from the user, never from you.** Testimonials, customer names and logos, results, uptime, prices and guarantees appear only when supplied. Missing proof: leave the slot out, or mark it `[Needs proof: …]` in the warning color so it cannot ship by accident. A believable invented claim is worse than an obvious placeholder.
+
+**Illustrative content looks real.** Sample data in product shots and UI (table rows, chart values, people in a demo inbox):
+- No placeholder people or brands (John Doe, Jane Doe, Acme, Nexus, SmartFlow); distinct avatars; varied dates.
+- No round sample numbers (99.99%, 50%, 1,000+ customers, 10x). In a claim, a round number signals it was made up: use the supplied figure or drop it.
+- No buzzwords: elevate, seamless, effortless, supercharge, streamline, revolutionize, unlock, empower, unleash, next-gen, game-changer, delve, tapestry, "in the world of".
+- Sentence-case headings. Errors say what happened and what to do, never "Oops!". No exclamation marks in success messages.
+
+**Ship floor.** Head tags apply to a full HTML document, not a component or fragment. Every full document has `<title>`, a meta description, a favicon (an inline SVG data URI from the brand mark when no file is supplied), and on a page meant for sharing `og:title` and `og:description`; `og:image` only from a supplied URL. A "Skip to content" link is the first focusable element, hidden until focused, targeting the `id` of `<main>`. Landmarks (`<header>`, `<nav>`, `<main>`, `<footer>`); alt text on meaningful images, `alt=""` on decorative ones. A marketing footer links privacy and terms at supplied URLs. Required fields and email format validate inline, wired with `aria-invalid` and `aria-describedby`. Every link goes somewhere real: no `href="#"`, no made-up path; a control with no target renders disabled. Navigation to other pages marks the current one `aria-current="page"`; in-page section links do not.
+
+**Values come from the scale.** Every font size, space and radius is a step of the system's scale (DESIGN.md, else the spec), chosen by role. An off-scale value snaps to the nearest step that keeps its role, usually the one below, with that step's line height; never below 16px body text or a 44px touch target, and an approved mockup's measured values stand. A shape nested in another with a gap under 32px gets inner radius = outer − gap when that exceeds 2px.
+
+**Line breaks.** Browser-wrapped headings get `text-wrap: balance`, body `text-wrap: pretty`. Break display lines where the thought breaks; never one word alone on the last line.
+
+**Motion.** These govern the one authored motion moment and state transitions; they license no more motion.
+- Content is visible by default; script adds the hidden start state only after the observer is set up (`.motion-ready` on `<html>`), so a script failure or a crawler sees everything and the CTA never waits on an animation.
+- Reveal with `IntersectionObserver` (`threshold: 0`, bottom `rootMargin` about −20%, once). Never an unthrottled `window.addEventListener('scroll')`; scroll-linked effects read scroll inside `requestAnimationFrame`.
+- Animate `transform`, `opacity`, `filter`, `clip-path` only; blur 14px or less.
+- No object grows from `scale(0)` (start at 0.85 or more); only lines and bars grow from zero. A data bar fills as a left-to-right `clip-path` wipe, 0.7–1.1s.
+- Stagger in reading order inside one section (30–80ms), never across the page.
+- Curves are named tokens: DESIGN.md's, else `cubic-bezier(0.16, 1, 0.3, 1)` for entrances and `cubic-bezier(0.65, 0, 0.35, 1)` for wipes; `linear` only for constant motion.
+- Loops pause off-screen. `prefers-reduced-motion: reduce` shows the final state. Full-height sections use `min-height: 100dvh`, not `100vh`.
+
+### Craft check
+
+At Step 4, after the Slop Gate and before the screenshots, run the free checker on a vanilla HTML output (for framework output, say it is unavailable):
+
+```bash
+{{BIN_DIR}}/gstack-design-craft <path-to-finalized.html>
+```
+
+It prints one `CRAFT_CHECK: <id> <count> <fix>` line per tell (the realism rules above, `[Needs proof:` markers, scroll listeners, `100vh`, `transition: all`, `href="#"`, `alert()`, missing alt; on a full document a missing title, description, favicon, working skip link, or `aria-current` on route navigation), then `CRAFT_RESULT: clean` or `CRAFT_RESULT: <n> issue(s)`. No `CRAFT_RESULT` line: say the check was unavailable and continue. Same contract as the Slop Gate: one surgical fix pass, one rerun, the rest listed as accepted-with-reason (a real customer named Nexus, a supplied 50% figure). Rerun once when the refinement loop exits; report only new lines.
+
+**Before this can ship.** Open the Step 4 hand-off, and repeat before Step 5's "Copy to project": every `[Needs proof: …]` marker, the dependencies the page cannot own (legal URLs, form endpoint, og:image, indexing, analytics and cookie consent where required), the accepted exceptions, and every claim the check made you change.
+
+Adapted from Elaya's [ai-design-skills](https://github.com/elayadesign/ai-design-skills) and [redesign-skill](https://github.com/elayadesign/redesign-skill) (MIT, © 2026 Elaya), rewritten for gstack. See NOTICE.md.

design-html/sections/manifest.json

@@ -16,6 +16,12 @@
       "title": "Pretext wiring patterns + API reference",
       "trigger": "writing the finalized HTML in Step 3 — the Pretext wiring patterns and API cheatsheet are the required reference for all text-layout code"
     },
+    {
+      "id": "craft",
+      "file": "craft.md",
+      "title": "Craft: Persuade pages and the craft floor",
+      "trigger": "writing the Step 1 implementation spec — the Persuade intake and archetype, and the craft floor every page ships with"
+    },
     {
       "id": "detector-install-offer",
       "file": "detector-install-offer.md",

design-shotgun/SKILL.md

@@ -634,6 +634,9 @@ Draw on DESIGN.md, taste memory, and the user's request to make each concept dis
 exists, it decides what varies: keep its fonts and palette and vary layout and
 composition. Without one, vary font family, color palette, and layout approach. If two
 variants look like siblings, rework the weaker one with a deliberately different direction.
+For a landing, pricing or campaign screen, also vary the page archetype (classic hero,
+long-form story, minimal conversion, comparison), keep the one primary action identical,
+and caption each variant with its section outline.
 
 Concrete test: if someone could swap the headline text between two variants without
 noticing, they're too similar. Variants should feel like separate design teams made

design-shotgun/SKILL.md.tmpl

@@ -219,6 +219,9 @@ Draw on DESIGN.md, taste memory, and the user's request to make each concept dis
 exists, it decides what varies: keep its fonts and palette and vary layout and
 composition. Without one, vary font family, color palette, and layout approach. If two
 variants look like siblings, rework the weaker one with a deliberately different direction.
+For a landing, pricing or campaign screen, also vary the page archetype (classic hero,
+long-form story, minimal conversion, comparison), keep the one primary action identical,
+and caption each variant with its section outline.
 
 Concrete test: if someone could swap the headline text between two variants without
 noticing, they're too similar. Variants should feel like separate design teams made

lib/design-craft.ts

@@ -0,0 +1,110 @@
+/**
+ * Free craft check for a generated page (design-html Step 4). Counts the tells
+ * that make a page read as generated or unfinished: placeholder content, AI
+ * cliché copy, scroll listeners, 100vh, dead links, and the missing ship-floor
+ * tags. Copy checks read the visible text only (script, style and tags
+ * removed); code checks read the source. Head checks run only on a full
+ * document. Adapted from Elaya's landing-page-design and redesign-skill
+ * (MIT); see NOTICE.md. Buzzwords report under the catalog's
+ * `marketing-buzzword` id, and BUZZWORDS covers every word its heuristic
+ * names (test-enforced), so the two lists cannot drift.
+ */
+
+/** The catalog's marketing-buzzword words plus Elaya's AI cliché words. */
+export const BUZZWORDS: readonly string[] = [
+  'seamless', 'effortless', 'supercharge', 'streamline', 'revolutionize', 'unlock', 'empower', 'elevate',
+  'unleash', 'next-gen', 'game-changer', 'delve', 'tapestry', 'in the world of',
+];
+
+function buzzwordPattern(): RegExp {
+  const stem = (w: string) => w.replace(/-/g, '[- ]').replace(/e$/, '');
+  return new RegExp(`\\b(?:${BUZZWORDS.map(stem).join('|')})(?:e|es|ed|ing|ly|er|ers|s|d|eration)?\\b`, 'gi');
+}
+
+export interface CraftFinding {
+  id: string;
+  count: number;
+  fix: string;
+}
+
+interface CountRule {
+  id: string;
+  on: 'source' | 'text' | 'script';
+  pattern: RegExp;
+  fix: string;
+}
+
+interface PresenceRule {
+  id: string;
+  pattern: RegExp;
+  fix: string;
+}
+
+const COUNT_RULES: CountRule[] = [
+  { id: 'placeholder-name', on: 'text', pattern: /\b(?:John|Jane) Doe\b/gi, fix: 'use a supplied name, or one that fits the product in sample data' },
+  { id: 'placeholder-brand', on: 'text', pattern: /\b(?:Acme|Nexus|SmartFlow)\b/g, fix: 'use a supplied customer, or a fitting name in sample data' },
+  { id: 'lorem', on: 'text', pattern: /\blorem ipsum\b/gi, fix: 'write real copy' },
+  { id: 'marketing-buzzword', on: 'text', pattern: buzzwordPattern(), fix: 'say the specific thing' },
+  { id: 'round-number', on: 'text', pattern: /(?<![\d.])(?:99\.9+|[1-9]0|100)%\s+(?:uptime|faster|more|less|fewer|better|growth|increase|reduction|of (?:teams|users|customers|companies))\b|\b(?:\d+0|\d+(?:,\d{3})+)\+?\s+(?:happy |active )?(?:users|customers|teams|companies)\b|(?<![\d.])10x\s+(?:faster|more|better)\b/gi, fix: 'use the supplied figure, or drop the claim' },
+  { id: 'needs-proof', on: 'text', pattern: /\[Needs proof:/g, fix: 'supply the proof or remove the claim before publishing' },
+  { id: 'oops', on: 'text', pattern: /\boops\b/gi, fix: 'say what happened and what to do' },
+  { id: 'scroll-listener', on: 'source', pattern: /addEventListener\(\s*['"`]scroll['"`]/g, fix: 'use IntersectionObserver; read scroll inside requestAnimationFrame for a scroll-linked effect' },
+  { id: 'transition-all', on: 'source', pattern: /\btransition\s*:\s*all\b|\btransition-all\b/g, fix: 'list the transitioned properties' },
+  { id: 'dead-link', on: 'source', pattern: /\bhref\s*=\s*["']#["']/g, fix: 'link it, or render the control disabled' },
+  { id: 'alert', on: 'script', pattern: /(?<![\w.$])(?:window\.)?alert\s*\(/g, fix: 'show the message inline' },
+  { id: 'img-no-alt', on: 'source', pattern: /<img\b(?![^>]*(?<![\w-])alt\s*=)[^>]*>/gi, fix: 'add alt text (alt="" when decorative)' },
+];
+
+const HEAD_RULES: PresenceRule[] = [
+  { id: 'missing-title', pattern: /<title>\s*[^<\s][^<]*<\/title>/i, fix: 'add a <title>' },
+  { id: 'missing-meta-description', pattern: /<meta\b[^>]*\bname\s*=\s*["']description["']/i, fix: 'add <meta name="description">' },
+  { id: 'missing-favicon', pattern: /<link\b[^>]*\brel\s*=\s*["'][^"']*\bicon\b/i, fix: 'add a favicon <link rel="icon">' },
+];
+
+export function visibleText(html: string): string {
+  return html
+    .replace(/<!--[\s\S]*?-->/g, ' ')
+    .replace(/<(script|style|template)\b[\s\S]*?<\/\1>/gi, ' ')
+    .replace(/<[^>]+>/g, ' ')
+    .replace(/&#(\d+);/g, (_, n) => String.fromCodePoint(Number(n)))
+    .replace(/&#x([0-9a-f]+);/gi, (_, n) => String.fromCodePoint(parseInt(n, 16)))
+    .replace(/&(nbsp|lt|gt|quot|apos|amp);/g, (_, e) => ({ nbsp: ' ', lt: '<', gt: '>', quot: '"', apos: "'", amp: '&' })[e as string] as string)
+    .replace(/\s+/g, ' ');
+}
+
+export function checkCraft(html: string): CraftFinding[] {
+  const text = visibleText(html);
+  const scripts = [...html.matchAll(/<script\b[^>]*>([\s\S]*?)<\/script>/gi), ...html.matchAll(/\bon[a-z]+\s*=\s*"([^"]*)"/gi)].map(m => m[1]).join('\n');
+  const findings: CraftFinding[] = [];
+  for (const rule of COUNT_RULES) {
+    const count = (rule.on === 'text' ? text : rule.on === 'script' ? scripts : html).match(rule.pattern)?.length ?? 0;
+    if (count > 0) findings.push({ id: rule.id, count, fix: rule.fix });
+  }
+  const vh = [...html.matchAll(/\b((?:min-|max-)?height)\s*:\s*100vh\b(\s*!important)?([^}]*)/g)]
+    .filter(m => m[2] || !new RegExp(`(?<![\\w-])${m[1]}\\s*:\\s*100[dsl]vh\\b`).test(m[3])).length
+    + (html.match(/\b(?:min-)?h-screen\b/g)?.length ?? 0);
+  if (vh > 0) findings.push({ id: 'vh-height', count: vh, fix: 'use min-height: 100dvh (a 100vh line followed by a 100dvh fallback is fine)' });
+  if (!/<head\b/i.test(html)) return findings;
+  for (const rule of HEAD_RULES) {
+    if (!rule.pattern.test(html)) findings.push({ id: rule.id, count: 1, fix: rule.fix });
+  }
+  const body = html.slice(html.search(/<body\b/i) + 1);
+  const firstFocusable = body.match(/<(?:a\b[^>]*\bhref|button|input|select|textarea)\b[^>]*>/i)?.[0] ?? '';
+  const skipTarget = firstFocusable.match(/^<a\b[^>]*\bhref\s*=\s*["']#([^"']+)["']/i)?.[1];
+  const main = html.match(/<main\b[^>]*(?<![\w-])id\s*=\s*["']([^"']+)["']/i)?.[1]
+    ?? html.match(/<[a-z]+\b[^>]*\brole\s*=\s*["']main["'][^>]*(?<![\w-])id\s*=\s*["']([^"']+)["']/i)?.[1];
+  if (!skipTarget || skipTarget !== main) {
+    findings.push({ id: 'missing-skip-link', count: 1, fix: 'add a "Skip to content" link first in <body> whose #id exists on <main>' });
+  }
+  const routeLinks = html.match(/<nav\b[\s\S]*?<\/nav>/gi)?.join('').match(/<a\b[^>]*\bhref\s*=\s*["'](?!#)[^"']+["']/gi)?.length ?? 0;
+  if (routeLinks >= 2 && !/\baria-current\s*=\s*["']page["']/i.test(html)) {
+    findings.push({ id: 'missing-aria-current', count: 1, fix: 'mark the current nav link aria-current="page"' });
+  }
+  return findings;
+}
+
+export function formatCraft(findings: CraftFinding[]): string {
+  const lines = findings.map(f => `CRAFT_CHECK: ${f.id} ${f.count} ${f.fix}`);
+  lines.push(findings.length === 0 ? 'CRAFT_RESULT: clean' : `CRAFT_RESULT: ${findings.length} issue(s)`);
+  return lines.join('\n');
+}

licenses/MIT-elaya.txt

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Elaya
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

package.json

@@ -1,6 +1,6 @@
 {
   "name": "gstack",
-  "version": "1.91.220",
+  "version": "1.91.226",
   "description": "Garry's Stack — Claude Code skills + fast headless browser. One repo, one install, entire AI engineering workflow.",
   "license": "MIT",
   "type": "module",

scripts/harness-version.json

@@ -29,7 +29,7 @@
     "test/helpers/eval-budgets.ts": {"blob":"667d0e09b20fe5d7366f57b77b5fb9b04baa37ef","since":6},
     "test/helpers/eval-store.ts": {"blob":"a50ea2c28f508ad9f35ea9353dd2f502ee0dcfc2","since":6,"nonBehavioral":"credential rejections are classed infra instead of assertion: no case's pass or fail changes; only failed trials whose provider rejected the run's key move class, so no rerun spends on a rejected key"},
     "test/helpers/hermetic-env.ts": {"blob":"cc072895461dc52ea317d1beff72121342420d30","since":6},
-    "test/helpers/llm-judge.ts": {"blob":"d0bc6bbbe9f478603d11a1ece95e66c812ad34e6","since":6},
+    "test/helpers/llm-judge.ts": {"blob":"a3177665b662932ce485ccc976c2b36a1531a0bb","since":6,"nonBehavioral":"callJudge rewrites only the message of a JSON parse failure (names the judge output, stop reason, length and the retained llm-judge-response-parse-error event); it still throws a SyntaxError before any score is parsed, the request is unchanged, and no case's pass or fail changes"},
     "test/helpers/paid-test-set.ts": {"blob":"ff9bfe43c43ec2bdd11114f7b8ec7c5736116f75","since":6},
     "test/helpers/pty-screen.ts": {"blob":"4f77569f9ce11f9cee9f347216e468cd7c956820","since":6},
     "test/helpers/pty/auq.ts": {"blob":"16e5ea2bb5b9f28f44808a1a09024dbd3e6f28c4","since":6},

scripts/psychographic-signals.ts

@@ -223,6 +223,89 @@ export const SIGNAL_MAP: Record<string, Record<string, DimensionDelta[]>> = {
   },
 };
 
+/**
+ * Label cues: option key → leading words of UI labels that mean that key.
+ *
+ * The AskUserQuestion hook only sees the label the user clicked ("A) Fix it
+ * now (recommended)"), never a stable key, while SIGNAL_MAP is keyed by the
+ * registry's option keys (#2623). resolveOptionKey() maps a label back to a
+ * key when bin/gstack-question-log records a hook event. Every key also
+ * matches its own words ("fix-now" → "fix now"), so only phrasings beyond
+ * that need an entry here. Cues are
+ * lowercase words separated by single spaces and match at the start of the
+ * normalized label; candidates are limited to the question's own registered
+ * options, and the longest cue wins.
+ */
+const OPTION_LABEL_CUES: Record<string, readonly string[]> = {
+  accept: ['yes', 'approve', 'agree', 'add', 'include', 'apply', 'adopt', 'enable', 'go ahead', 'do it', 'proceed', 'run', 'create', 'merge', 'roll back', 'migrate', 'open', 'promote', 'reorganize', 'polish'],
+  reject: ['no', 'disagree', 'decline', 'cancel', 'abort', 'dont', 'do not', 'keep current', 'keep as is', 'keep it as is'],
+  skip: ['no', 'no thanks', 'not now', 'dont', 'do not', 'pass', 'drop', 'ignore', 'leave it', 'ship as is', 'ship it as is'],
+  defer: ['later', 'fix later', 'add to todos', 'add to todo', 'add a todo', 'todo', 'postpone', 'backlog', 'park', 'track it'],
+  'fix-now': ['fix', 'fix it', 'fix them', 'fix before shipping'],
+  'fix-anyway': ['fix', 'fix it'],
+  investigate: ['dig in', 'look into', 'debug', 'diagnose', 'root cause'],
+  'ack-and-ship': ['ack', 'acknowledge', 'ship', 'ship anyway', 'ship it'],
+  'accept-risk': ['accept', 'accept the risk', 'live with it'],
+  'false-positive': ['not a bug', 'mark false positive', 'mark as false positive', 'mark it false positive'],
+  'add-test': ['add', 'add tests', 'add the test', 'write', 'write the test', 'test it'],
+  expand: ['scope expansion'],
+  selective: ['cherry pick', 'selective expansion'],
+  hold: ['keep', 'keep scope', 'hold scope'],
+  reduce: ['cut', 'trim', 'shrink', 'scope down', 'scope reduction'],
+  revise: ['edit', 'change', 'rework'],
+  restart: ['start over', 'redo'],
+  'build-now': ['build', 'build it now', 'include in this pr'],
+  'keep-local': ['keep'],
+  'keep-vendored': ['keep'],
+  'yes-upgrade': ['yes', 'upgrade'],
+  'always-auto': ['always'],
+  'not-now': ['later', 'snooze'],
+  'never-ask': ['never', 'disable'],
+};
+
+function normalizeOptionLabel(label: string): string {
+  return label
+    .toLowerCase()
+    .replace(/\s*\(recommended\)\s*$/, '')
+    .replace(/['\u2019]/g, '')
+    .replace(/^\s*(?:option\s+)?[a-z0-9]\s*(?:[).:]|-(?=\s))\s+/, '')
+    .replace(/[^a-z0-9]+/g, ' ')
+    .trim();
+}
+
+/**
+ * Map a chosen UI label to the registered option key of its question.
+ *
+ * Returns the key when exactly one of the question's registered options
+ * matches with the longest cue; undefined for unregistered ids, questions
+ * without options, unmatched labels and ties, so an unclear answer stays
+ * unmapped instead of nudging the wrong dimension.
+ */
+export function resolveOptionKey(question_id: string, choice: string): string | undefined {
+  const options = (QUESTIONS as Record<string, QuestionDef>)[question_id]?.options as string[] | undefined;
+  if (!options?.length) return undefined;
+  if (options.includes(choice)) return choice;
+  const label = normalizeOptionLabel(choice);
+  if (!label) return undefined;
+  let best: string | undefined;
+  let bestWords = 0;
+  let tie = false;
+  for (const key of options) {
+    for (const cue of [key.replace(/-/g, ' '), ...(OPTION_LABEL_CUES[key] ?? [])]) {
+      if (label !== cue && !label.startsWith(cue + ' ')) continue;
+      const words = cue.split(' ').length;
+      if (words > bestWords) {
+        best = key;
+        bestWords = words;
+        tie = false;
+      } else if (words === bestWords && best !== key) {
+        tie = true;
+      }
+    }
+  }
+  return tie ? undefined : best;
+}
+
 /**
  * Apply a user choice for a question to the running dimension totals.
  *

test/context-restore-fragment-verdict.test.ts

@@ -0,0 +1,121 @@
+import { describe, test, expect } from 'bun:test';
+import * as fs from 'fs';
+import * as path from 'path';
+import { parseNDJSON } from './helpers/session-runner';
+import { fragmentRestoreVerdict } from './helpers/context-restore-fragment';
+
+const MARKERS = { restored: 'FRAGMATCH_PAYMENTS_BUILD', others: ['FRAGMATCH_ALPHA_BUILD', 'FRAGMATCH_OMEGA_BUILD'] };
+const BODIES: Record<string, string> = {
+  '20260101-120000-alpha-feature.md': 'branch: feat/alpha\n## Working on: alpha feature\n### Summary\nAlpha content FRAGMATCH_ALPHA_BUILD\n',
+  '20260202-120000-middle-payments.md': 'branch: feat/payments\n## Working on: middle payments\n### Summary\nPayments content FRAGMATCH_PAYMENTS_BUILD\n',
+  '20260303-120000-omega-release.md': 'branch: feat/omega\n## Working on: omega release\n### Summary\nOmega content FRAGMATCH_OMEGA_BUILD\n',
+};
+const ALL_FILES = Object.keys(BODIES);
+
+const captures: { sessions: Array<{ session_id: string; ndjson: string[] }> } = JSON.parse(
+  fs.readFileSync(path.join(import.meta.dir, 'fixtures', 'context-restore-fragment-match-captures.json'), 'utf8'),
+);
+
+function replay(lines: string[]) {
+  const parsed = parseNDJSON(lines);
+  return { ...parsed, output: parsed.resultLine?.result || '' };
+}
+
+function toolOutput(run: ReturnType<typeof replay>): string {
+  return run.toolCalls.map((call) => call.output).join('\n');
+}
+
+function isReportLine(line: string): boolean {
+  const event = JSON.parse(line);
+  return event.type === 'result'
+    || (event.type === 'assistant' && event.message.content.some((block: any) => block.type === 'text'));
+}
+
+function rewriteReport(lines: string[], rewrite: (line: string) => string): string[] {
+  return lines.map((line) => (isReportLine(line) ? rewrite(line) : line));
+}
+
+function synthetic(files: string[], report: string): string[] {
+  const lines: string[] = [
+    JSON.stringify({ type: 'assistant', parent_tool_use_id: null, message: { content: [{ type: 'tool_use', id: 'skill', name: 'Skill', input: { skill: 'context-restore', args: 'payments' } }] } }),
+    JSON.stringify({ type: 'user', parent_tool_use_id: null, message: { content: [{ type: 'tool_result', tool_use_id: 'skill', content: 'Launching skill: context-restore' }] } }),
+  ];
+  files.forEach((file, i) => {
+    lines.push(JSON.stringify({ type: 'assistant', parent_tool_use_id: null, message: { content: [{ type: 'tool_use', id: `read-${i}`, name: 'Read', input: { file_path: `.gstack-home/projects/p/checkpoints/${file}` } }] } }));
+    lines.push(JSON.stringify({ type: 'user', parent_tool_use_id: null, message: { content: [{ type: 'tool_result', tool_use_id: `read-${i}`, content: BODIES[file] }] } }));
+  });
+  if (report) {
+    lines.push(JSON.stringify({ type: 'assistant', parent_tool_use_id: null, message: { content: [{ type: 'text', text: report }] } }));
+  }
+  lines.push(JSON.stringify({ type: 'result', subtype: report ? 'success' : 'error_max_turns', result: report }));
+  return lines;
+}
+
+function resumeReport(file: string, ...summaries: string[]): string {
+  return `I loaded \`${file}\`.\n\nRESUMING CONTEXT\n### Summary\n${summaries.join('\n')}\n`;
+}
+
+describe('context-restore-fragment-match verdict', () => {
+  test('captured runs hold three sessions that each printed another checkpoint body', () => {
+    expect(captures.sessions).toHaveLength(3);
+    for (const session of captures.sessions) {
+      const run = replay(session.ndjson);
+      expect(run.toolCalls.map((call) => call.tool)).toContain('Skill');
+      expect(toolOutput(run)).toContain('FRAGMATCH_PAYMENTS_BUILD');
+      expect(MARKERS.others.some((marker) => toolOutput(run).includes(marker))).toBe(true);
+    }
+  });
+
+  test.each(captures.sessions.map((s) => [s.session_id, s.ndjson] as const))(
+    'captured run %s: inspected every checkpoint, restored payments -> pass',
+    (_id, ndjson) => {
+      expect(fragmentRestoreVerdict(replay(ndjson), MARKERS)).toEqual({ loadedTarget: true, didNotLoadOthers: true });
+    },
+  );
+
+  test('synthetic: prints every checkpoint, reports payments -> pass', () => {
+    const run = replay(synthetic(ALL_FILES, resumeReport(ALL_FILES[1], 'Payments content FRAGMATCH_PAYMENTS_BUILD')));
+    expect(fragmentRestoreVerdict(run, MARKERS)).toEqual({ loadedTarget: true, didNotLoadOthers: true });
+  });
+
+  test.each([
+    ['alpha', 'FRAGMATCH_ALPHA_BUILD'],
+    ['omega', 'FRAGMATCH_OMEGA_BUILD'],
+  ])('captured run rewritten to resume %s -> fail', (_name, marker) => {
+    for (const session of captures.sessions) {
+      const run = replay(rewriteReport(session.ndjson, (line) => line.replaceAll('FRAGMATCH_PAYMENTS_BUILD', marker)));
+      expect(toolOutput(run)).toContain('FRAGMATCH_PAYMENTS_BUILD');
+      expect(fragmentRestoreVerdict(run, MARKERS)).toEqual({ loadedTarget: false, didNotLoadOthers: false });
+    }
+  });
+
+  test.each([
+    ['alpha', ALL_FILES[0], 'Alpha content FRAGMATCH_ALPHA_BUILD'],
+    ['omega', ALL_FILES[2], 'Omega content FRAGMATCH_OMEGA_BUILD'],
+  ])('synthetic: reads only %s and resumes it -> fail', (_name, file, summary) => {
+    const run = replay(synthetic([file], resumeReport(file, summary)));
+    expect(fragmentRestoreVerdict(run, MARKERS)).toEqual({ loadedTarget: false, didNotLoadOthers: false });
+  });
+
+  test('captured run with omega content mixed into the resume summary -> fail', () => {
+    for (const session of captures.sessions) {
+      const mixed = rewriteReport(session.ndjson, (line) => line.replaceAll(
+        'Payments content FRAGMATCH_PAYMENTS_BUILD', 'Payments content FRAGMATCH_PAYMENTS_BUILD\\nOmega content FRAGMATCH_OMEGA_BUILD'));
+      expect(fragmentRestoreVerdict(replay(mixed), MARKERS)).toEqual({ loadedTarget: true, didNotLoadOthers: false });
+    }
+  });
+
+  test('synthetic: resume summary mixes alpha and payments -> fail', () => {
+    const run = replay(synthetic(ALL_FILES, resumeReport(ALL_FILES[1],
+      'Payments content FRAGMATCH_PAYMENTS_BUILD', 'Alpha content FRAGMATCH_ALPHA_BUILD')));
+    expect(fragmentRestoreVerdict(run, MARKERS)).toEqual({ loadedTarget: true, didNotLoadOthers: false });
+  });
+
+  test('payments body only in tool output, no report of it -> fail', () => {
+    const noReport = replay(synthetic([ALL_FILES[1]], ''));
+    expect(toolOutput(noReport)).toContain('FRAGMATCH_PAYMENTS_BUILD');
+    expect(fragmentRestoreVerdict(noReport, MARKERS).loadedTarget).toBe(false);
+    const nameOnly = replay(synthetic(ALL_FILES, `I loaded \`${ALL_FILES[1]}\`.`));
+    expect(fragmentRestoreVerdict(nameOnly, MARKERS).loadedTarget).toBe(false);
+  });
+});

test/context-restore-head.test.ts

@@ -0,0 +1,202 @@
+/**
+ * /context-save records the commit a checkpoint describes, and /context-restore
+ * notices when commits landed after it, end to end.
+ *
+ * Runs the REAL save-path, stamp and head-check bash blocks extracted from the
+ * generated context-save and context-restore SKILL.md files (with
+ * ~/.claude/skills/gstack pointed at this checkout) inside fixture repos. The bug
+ * these pin: a checkpoint saved before work landed came back as the current plan,
+ * with nothing saying HEAD had moved, so restore offered to redo finished steps.
+ *
+ * Value: protects=save stamps head: and restore reports unchanged, moved, diverged, untracked and unknown commits; fails_when=the stamp drops head: or restore stops comparing it with HEAD; why_new=no test compares a checkpoint with commits made after it; seam=none; no_claim=fixture repos and a hand-written checkpoint body stand in for a live /context-save, so green does not prove the agent follows the prompt
+ */
+import { describe, test, expect, beforeEach, afterEach } from 'bun:test';
+import { spawnSync } from 'child_process';
+import * as fs from 'fs';
+import * as os from 'os';
+import * as path from 'path';
+
+const ROOT = path.resolve(import.meta.dir, '..');
+// Forward slashes keep the path intact inside bash on Windows (Git Bash); a no-op elsewhere.
+const BASH_ROOT = ROOT.split(path.sep).join('/');
+
+function bashBlockAfter(skillFile: string, heading: string, contains?: string): string {
+  const md = fs.readFileSync(path.join(ROOT, skillFile), 'utf-8');
+  const start = md.indexOf(heading);
+  if (start < 0) throw new Error(`${skillFile}: heading not found: ${heading}`);
+  const re = /```bash\n([\s\S]*?)```/g;
+  re.lastIndex = start;
+  for (let m = re.exec(md); m; m = re.exec(md)) {
+    if (!contains || m[1].includes(contains)) return m[1].replaceAll('~/.claude/skills/gstack', BASH_ROOT);
+  }
+  throw new Error(`${skillFile}: no bash block after ${heading}`);
+}
+
+const SAVE_PATH_BLOCK = bashBlockAfter('context-save/SKILL.md', '### Step 4: Write saved-context file').replace(
+  'TIMESTAMP=$(date +%Y%m%d-%H%M%S)',
+  'TIMESTAMP="$TEST_TIMESTAMP"',
+);
+const STAMP_BLOCK = bashBlockAfter('context-save/SKILL.md', '### Step 4: Write saved-context file', '--stamp-checkpoint');
+const HEAD_BLOCK = bashBlockAfter('context-restore/SKILL.md', '**Check whether the branch moved since the save.**', 'SAVED_HEAD');
+
+let tmp: string;
+let home: string;
+
+function env(extra: Record<string, string> = {}): Record<string, string> {
+  const out: Record<string, string> = {};
+  for (const [k, v] of Object.entries(process.env)) if (v !== undefined) out[k] = v;
+  delete out.GSTACK_PROJECT_SLUG;
+  delete out.GSTACK_STATE_ROOT;
+  delete out.GSTACK_STATE_DIR;
+  delete out.CURRENT_BRANCH;
+  delete out.CLAUDE_PROJECT_DIR;
+  return { ...out, HOME: tmp, GSTACK_HOME: home, ...extra };
+}
+
+function sh(script: string, cwd: string, extra: Record<string, string> = {}) {
+  const r = spawnSync('bash', ['-c', script], { cwd, env: env(extra), encoding: 'utf-8', timeout: 30_000 });
+  return { stdout: r.stdout ?? '', stderr: r.stderr ?? '', status: r.status ?? -1 };
+}
+
+function git(cwd: string, ...args: string[]): string {
+  const r = spawnSync('git', args, { cwd, encoding: 'utf-8', timeout: 30_000 });
+  if (r.status !== 0) throw new Error(`git ${args.join(' ')}: ${r.stderr}`);
+  return r.stdout.trim();
+}
+
+function commit(cwd: string, subject: string) {
+  fs.appendFileSync(path.join(cwd, 'app.txt'), `${subject}\n`);
+  git(cwd, 'add', 'app.txt');
+  git(cwd, 'commit', '-q', '-m', subject);
+}
+
+function repo(): string {
+  const dir = path.join(tmp, 'app');
+  fs.mkdirSync(dir, { recursive: true });
+  git(dir, 'init', '-q', '-b', 'main');
+  git(dir, 'remote', 'add', 'origin', 'https://github.com/example/head-check.git');
+  git(dir, 'config', 'user.name', 'gstack test');
+  git(dir, 'config', 'user.email', 'gstack-test@example.invalid');
+  commit(dir, 'baseline');
+  return dir;
+}
+
+/** The context-save flow: path block, the agent's Write, then the stamp block. */
+function save(cwd: string): string {
+  const paths = sh(SAVE_PATH_BLOCK, cwd, { TITLE_RAW: 'login flow', TEST_TIMESTAMP: '20261020-090000' });
+  expect(paths.status).toBe(0);
+  const file = paths.stdout.match(/^FILE=(.+)$/m)![1];
+  fs.writeFileSync(file, '---\nstatus: in-progress\nbranch: main\n---\n\n## Working on: login flow\n');
+  const stamp = sh(STAMP_BLOCK.replace('FILE="<the FILE path printed above>"', `FILE=${JSON.stringify(file)}`), cwd);
+  expect(stamp.stdout).toContain('STAMPED');
+  return file;
+}
+
+function headCheck(cwd: string, file: string) {
+  const r = sh(HEAD_BLOCK.replace('FILE="<the FILE path printed above>"', `FILE=${JSON.stringify(file)}`), cwd);
+  expect(r.status).toBe(0);
+  return r.stdout;
+}
+
+beforeEach(() => {
+  tmp = fs.realpathSync(fs.mkdtempSync(path.join(os.tmpdir(), 'ctx-head-')));
+  home = path.join(tmp, '.gstack');
+});
+
+afterEach(() => {
+  fs.rmSync(tmp, { recursive: true, force: true });
+});
+
+describe('context-save records the commit a checkpoint describes', () => {
+  test('the stamp writes head: once, and re-stamping does not duplicate it', () => {
+    const app = repo();
+    const file = save(app);
+    const fm = fs.readFileSync(file, 'utf-8').split('---')[1];
+    expect(fm).toContain(`\nhead: ${git(app, 'rev-parse', 'HEAD')}\n`);
+    commit(app, 'later work');
+    const restamp = spawnSync('bash', [path.join(ROOT, 'bin', 'gstack-slug'), '--stamp-checkpoint', file], { cwd: app, env: env(), timeout: 30_000 });
+    expect(restamp.status).toBe(0);
+    expect(fs.readFileSync(file, 'utf-8').split('---')[1].match(/^head:.*$/gm)).toEqual([`head: ${git(app, 'rev-parse', 'HEAD')}`]);
+  });
+});
+
+describe('context-restore compares the checkpoint with HEAD', () => {
+  test('nothing to add when no commit landed after the save', () => {
+    const app = repo();
+    expect(headCheck(app, save(app)).trim()).toBe('HEAD_UNCHANGED');
+  });
+
+  test('commits that landed after the save are counted and listed', () => {
+    const app = repo();
+    const file = save(app);
+    commit(app, 'add the login route');
+    commit(app, 'add login tests');
+    const out = headCheck(app, file);
+    expect(out).toStartWith('HEAD_MOVED 2\n');
+    expect(out).toContain('add the login route');
+    expect(out).toContain('add login tests');
+  });
+
+  test('a checkpoint whose commit left the branch history is reported, not trusted', () => {
+    const app = repo();
+    commit(app, 'work in progress');
+    const file = save(app);
+    git(app, 'reset', '-q', '--hard', 'HEAD~1');
+    commit(app, 'rewritten work');
+    expect(headCheck(app, file)).toStartWith('HEAD_DIVERGED ');
+  });
+
+  test('a checkpoint resumed from another branch is reported, and the prompt names that cause', () => {
+    const app = repo();
+    git(app, 'checkout', '-q', '-b', 'feature');
+    commit(app, 'feature work');
+    const file = save(app);
+    git(app, 'checkout', '-q', 'main');
+    commit(app, 'unrelated main work');
+    expect(headCheck(app, file)).toStartWith('HEAD_DIVERGED ');
+    const restore = fs.readFileSync(path.join(ROOT, 'context-restore/SKILL.md'), 'utf-8');
+    expect(restore).toContain('(saved on another\n  branch, rebased, reset, or not fetched)');
+  });
+
+  test('checkpoints saved before head tracking, or with a malformed head, are never treated as current', () => {
+    const app = repo();
+    const legacy = path.join(tmp, 'legacy.md');
+    fs.writeFileSync(legacy, '---\nstatus: in-progress\nbranch: main\n---\n');
+    expect(headCheck(app, legacy).trim()).toBe('HEAD_UNTRACKED');
+    const bogus = path.join(tmp, 'bogus.md');
+    fs.writeFileSync(bogus, '---\nbranch: main\nhead: $(touch pwned)\n---\n');
+    expect(headCheck(app, bogus)).toStartWith('HEAD_UNKNOWN ');
+    expect(fs.existsSync(path.join(app, 'pwned'))).toBe(false);
+    const bodyOnly = path.join(tmp, 'body-only.md');
+    fs.writeFileSync(bodyOnly, `---\nstatus: in-progress\nbranch: main\n---\n\nhead: ${git(app, 'rev-parse', 'HEAD~0')}\n`);
+    expect(headCheck(app, bodyOnly).trim()).toBe('HEAD_UNTRACKED');
+  });
+
+  test('outside git, a save from a nested linked worktree is compared with that worktree, not the main checkout', () => {
+    const app = repo();
+    const wt = path.join(app, '.worktrees', 'side');
+    git(app, 'worktree', 'add', '-q', '-b', 'side', wt);
+    const file = save(wt);
+    expect(fs.readFileSync(file, 'utf-8')).toContain(`\nworktree: ${wt}\n`);
+    commit(app, 'main moves on');
+    commit(app, 'and again');
+    const outside = path.join(tmp, 'outside');
+    fs.mkdirSync(outside);
+    expect(headCheck(outside, file).trim()).toBe('HEAD_UNCHANGED');
+    commit(wt, 'worktree work');
+    expect(headCheck(outside, file)).toStartWith('HEAD_MOVED 1\n');
+  });
+
+  test('the prompt maps every status to what the user sees', () => {
+    const restore = fs.readFileSync(path.join(ROOT, 'context-restore/SKILL.md'), 'utf-8');
+    for (const status of ['HEAD_UNCHANGED', 'HEAD_NOT_CHECKED', 'HEAD_MOVED <n>', 'HEAD_DIVERGED', 'HEAD_UNKNOWN', 'HEAD_UNTRACKED']) {
+      expect(restore).toContain(`\`${status}\``);
+    }
+    expect(restore).toContain('`<n> commit(s) landed since this save:`');
+    expect(restore).toContain('under **Verify first**');
+    expect(restore).toContain('{commit-tracking lines, if any}');
+    expect(restore).toContain('When the head check printed `HEAD_MOVED`, `HEAD_DIVERGED`\nor `HEAD_UNKNOWN`, skip the rules below: every item goes under Verify first,\nwhatever its marker.');
+    const save = fs.readFileSync(path.join(ROOT, 'context-save/SKILL.md'), 'utf-8');
+    expect(save).toContain('`project_root:`, `worktree:` or `head:` yourself');
+  });
+});

test/design-craft.test.ts

@@ -0,0 +1,118 @@
+import { describe, expect, test } from 'bun:test';
+import * as fs from 'fs';
+import * as os from 'os';
+import * as path from 'path';
+import { BUZZWORDS, checkCraft, formatCraft, visibleText } from '../lib/design-craft';
+import { catalogEntry } from '../lib/design-catalog';
+
+const ROOT = path.resolve(import.meta.dir, '..');
+const BIN = path.join(ROOT, 'bin', 'gstack-design-craft');
+
+const DIRTY = `<!doctype html><html><head><style>.hero{height:100vh;transition: all .3s}</style></head><body>
+<nav><a href="#">Docs</a><a href="/">Home</a><a href="/pricing">Pricing</a></nav>
+<h1>Elevate your workflow seamlessly</h1><p>Trusted by Acme and John Doe. 99.99% uptime, 50% faster, 10,000+ happy users. [Needs proof: logo]</p>
+<img src="a.png"><p>Oops! Something broke. Lorem ipsum.</p>
+<script>window.addEventListener('scroll', () => {}); alert('hi'); showAlert(1);</script></body></html>`;
+
+const CLEAN = `<!doctype html><html lang="en"><head><title>Ledgerline: close the books in 2 days</title>
+<meta name="description" content="Close faster."><meta property="og:image" content="/og.png"><link rel="icon" href="/favicon.svg">
+<style>.hero{min-height:100vh;min-height:100dvh;transition: transform .4s var(--ease-out)} .w{width:100%}</style></head>
+<body><a href="#main">Skip to content</a><a href="#pricing">Pricing</a><nav><a href="/" aria-current="page">Home</a><a href="/pricing">Pricing</a></nav>
+<main id="main"><h1>Close the books in 2 days, not 9</h1><p>Teams at Halvorsen Freight cut month-end by 47.2%. Save 20% yearly. 100% open source. 30-day refund.</p><img src="x.png" alt=""></main>
+<script>const io = new IntersectionObserver(() => {}); window.showAlert = () => {};</script></body></html>`;
+
+function run(file?: string) {
+  const r = Bun.spawnSync([BIN, ...(file ? [file] : [])], { stdout: 'pipe', stderr: 'pipe', timeout: 30_000 });
+  return { code: r.exitCode, out: r.stdout.toString(), err: r.stderr.toString() };
+}
+
+describe('design-craft check', () => {
+  test('a page with every tell reports each one with its count', () => {
+    const counts = Object.fromEntries(checkCraft(DIRTY).map(f => [f.id, f.count]));
+    expect(counts).toEqual({
+      'placeholder-name': 1, 'placeholder-brand': 1, lorem: 1, 'marketing-buzzword': 2, 'round-number': 3, 'needs-proof': 1, oops: 1,
+      'scroll-listener': 1, 'transition-all': 1, 'dead-link': 1, alert: 1, 'img-no-alt': 1, 'vh-height': 1,
+      'missing-title': 1, 'missing-meta-description': 1, 'missing-favicon': 1,
+      'missing-skip-link': 1, 'missing-aria-current': 1,
+    });
+  });
+
+  test('a finished page is clean: CSS 100%, a 100vh line with its 100dvh fallback, in-page anchors, a showAlert helper and inlined Pretext are not flagged', () => {
+    expect(checkCraft(CLEAN)).toEqual([]);
+    const pretext = fs.readFileSync(path.join(ROOT, 'design-html', 'vendor', 'pretext.js'), 'utf-8');
+    expect(checkCraft(CLEAN.replace('</body>', `<script type="module">${pretext}</script></body>`))).toEqual([]);
+    expect(formatCraft([])).toBe('CRAFT_RESULT: clean');
+  });
+
+  test('BUZZWORDS covers every word the catalog marketing-buzzword heuristic names', () => {
+    const heuristic = catalogEntry('marketing-buzzword')?.heuristic ?? '';
+    const words = heuristic.replace(/^.*?for /, '').replace(/\.$/, '').split(/,\s*/);
+    expect(words.length).toBeGreaterThan(5);
+    for (const word of words) expect(BUZZWORDS).toContain(word);
+    expect(checkCraft('<p>Unlocking seamless, next gen workflows that elevate teams</p>').map(f => [f.id, f.count])).toEqual([['marketing-buzzword', 4]]);
+    expect(checkCraft('<p>Take the elevator to the elevation map; streamers sold here</p>')).toEqual([]);
+  });
+
+  test('the skip link must reach an existing id, and aria-current is required only for route navigation', () => {
+    const ids = (html: string) => checkCraft(html).map(f => f.id);
+    expect(ids(CLEAN.replace('id="main"', 'id="content"'))).toEqual(['missing-skip-link']);
+    const sectionNav = CLEAN.replace('<a href="/" aria-current="page">Home</a><a href="/pricing">Pricing</a>', '<a href="#features">Features</a><a href="#faq">FAQ</a>');
+    expect(ids(sectionNav)).toEqual([]);
+    expect(ids(CLEAN.replace(' aria-current="page"', ''))).toEqual(['missing-aria-current']);
+  });
+
+  test('attribute look-alikes, encoded markers and 100vh ordering are read correctly', () => {
+    const ids = (html: string) => checkCraft(html).map(f => f.id);
+    expect(ids('<img src="a.png" data-alt="x">')).toEqual(['img-no-alt']);
+    expect(ids('<p>&#91;Needs proof: logo]</p>')).toEqual(['needs-proof']);
+    expect(ids(CLEAN.replace('id="main"', 'data-id="main"'))).toEqual(['missing-skip-link']);
+    expect(ids('<style>.a{min-height:100dvh;min-height:100vh}</style>')).toEqual(['vh-height']);
+    expect(ids('<style>.a{min-height:100vh !important;min-height:100dvh}</style>')).toEqual(['vh-height']);
+  });
+
+  test('every word and name the craft section bans is one the checker flags', () => {
+    const section = fs.readFileSync(path.join(ROOT, 'design-html', 'sections', 'craft.md.tmpl'), 'utf-8');
+    const buzz = section.match(/No buzzwords: ([^\n]+)\.\n/)?.[1].replace(/"/g, '').split(/,\s*/) ?? [];
+    expect(buzz.length).toBeGreaterThan(10);
+    for (const word of buzz) expect(BUZZWORDS).toContain(word);
+    const names = section.match(/No placeholder people or brands \(([^)]+)\)/)?.[1].split(/,\s*/) ?? [];
+    expect(names.length).toBe(5);
+    for (const name of names) expect(checkCraft(`<p>${name}</p>`).map(f => f.id)).toEqual([/Doe/.test(name) ? 'placeholder-name' : 'placeholder-brand']);
+  });
+
+  test('skip link is judged by position and target; alert() counts only in script; svg text is copy', () => {
+    const ids = (html: string) => checkCraft(html).map(f => f.id);
+    expect(ids(CLEAN.replace('>Skip to content<', '>Zum Inhalt springen<'))).toEqual([]);
+    expect(ids(CLEAN.replace('<a href="#main">Skip to content</a>', ''))).toEqual(['missing-skip-link']);
+    expect(ids('<p>Set a price alert (free)</p>')).toEqual([]);
+    expect(ids('<button onclick="alert(1)">Go</button>')).toEqual(['alert']);
+    expect(ids('<svg><text>Acme</text></svg>')).toEqual(['placeholder-brand']);
+  });
+
+  test('copy rules read visible text only; head rules skip a fragment', () => {
+    expect(visibleText('<style>.a{content:"Acme"}</style><script>// John Doe</script><p>Halvorsen</p>')).toBe(' Halvorsen ');
+    expect(checkCraft('<section><h2>Pricing</h2><a href="/buy">Start free trial</a></section>')).toEqual([]);
+  });
+
+  test('the CLI prints CRAFT_CHECK lines and exits 1 on tells, 0 when clean, 2 on usage', () => {
+    const dir = fs.mkdtempSync(path.join(os.tmpdir(), 'design-craft-'));
+    try {
+      fs.writeFileSync(path.join(dir, 'dirty.html'), DIRTY);
+      fs.writeFileSync(path.join(dir, 'clean.html'), CLEAN);
+      const dirty = run(path.join(dir, 'dirty.html'));
+      expect(dirty.code).toBe(1);
+      expect(dirty.out).toContain('CRAFT_CHECK: scroll-listener 1 ');
+      expect(dirty.out.trim().split('\n').at(-1)).toBe('CRAFT_RESULT: 18 issue(s)');
+      const clean = run(path.join(dir, 'clean.html'));
+      expect(clean).toMatchObject({ code: 0, out: 'CRAFT_RESULT: clean\n' });
+      expect(run().code).toBe(2);
+      expect(run(path.join(dir, 'missing.html')).code).toBe(2);
+      fs.writeFileSync(path.join(dir, 'empty.html'), '  \n');
+      expect(run(path.join(dir, 'empty.html')).code).toBe(2);
+      fs.writeFileSync(path.join(dir, 'page.tsx'), 'export const P = () => <p>Acme</p>;');
+      expect(run(path.join(dir, 'page.tsx'))).toMatchObject({ code: 2, out: '' });
+    } finally {
+      fs.rmSync(dir, { recursive: true, force: true });
+    }
+  });
+});

test/fixtures/context-restore-fragment-match-captures.json

@@ -0,0 +1,51 @@
+{
+ "provenance": {
+  "case": "context-restore-fragment-match",
+  "run": "local paid baseline: scripts/test-paid-shards.ts --tier periodic --case context-restore-fragment-match --trials 3 (0/3, all on didNotLoadOthers)",
+  "head": "cb99df4a2fc45bc983e0a64f862f79fab0ce899a",
+  "claude_code": "2.1.292",
+  "source_sha256": "2d05a7257f7d559d60ec2e644178a3d65cd0799121260633c931d5b54755b495",
+  "projection": "Raw NDJSON lines of each session's assistant, user and result events, unchanged; system events, thinking-only assistant events and the synthetic user message carrying the extracted SKILL.md body dropped."
+ },
+ "sessions": [
+  {
+   "session_id": "d1147327-49e3-4371-ab54-784d135573ac",
+   "ndjson": [
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-5-5\",\"id\":\"msg_011CgP2Sp1UcGZ3JFpbUbxQg\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01HHxMjn7G6dQ36WFwvafm5W\",\"name\":\"Skill\",\"input\":{\"skill\":\"context-restore\",\"args\":\"payments\"},\"caller\":{\"type\":\"direct\"}}],\"container\":null,\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":6256,\"cache_read_input_tokens\":12042,\"cache_creation\":{\"ephemeral_5m_input_tokens\":6256,\"ephemeral_1h_input_tokens\":0},\"output_tokens\":16,\"service_tier\":\"standard\",\"inference_geo\":\"global\",\"speed\":\"standard\"},\"input_transformations\":[],\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"d1147327-49e3-4371-ab54-784d135573ac\",\"uuid\":\"691912ce-6742-4e21-8f4c-a3cfa15b4112\",\"timestamp\":\"2026-10-25T13:21:55.808Z\",\"request_id\":\"req_011CgP2Som6KWQBvN4W6W3B8\",\"wire_tool_inputs\":{\"toolu_01HHxMjn7G6dQ36WFwvafm5W\":{\"skill\":\"context-restore\",\"args\":\"payments\"}}}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"type\":\"tool_result\",\"tool_use_id\":\"toolu_01HHxMjn7G6dQ36WFwvafm5W\",\"content\":\"Launching skill: context-restore\"}]},\"parent_tool_use_id\":null,\"session_id\":\"d1147327-49e3-4371-ab54-784d135573ac\",\"uuid\":\"9e05cd18-bf0c-4d75-93d8-db0f817d8c87\",\"timestamp\":\"2026-10-25T13:21:55.840Z\",\"tool_use_result\":{\"success\":true,\"commandName\":\"context-restore\",\"allowedTools\":[\"Bash\",\"Read\",\"Glob\",\"Grep\",\"AskUserQuestion\"]}}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-5-5\",\"id\":\"msg_011CgP2SwJ2idXv6jZBjmMRS\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01EyWkH3v7Gu7cBDAgVzCiSu\",\"name\":\"Bash\",\"input\":{\"command\":\"cd /tmp/gstack-paid-shard-O15pOD/tmp/skill-e2e-ctx-fragment-svqfPa; ls bin; export GSTACK_HOME=\\\"$PWD/.gstack-home\\\"; B=./bin\\nGSTACK_STATE_ROOT=$($B/gstack-paths --get GSTACK_STATE_ROOT); echo ROOT=$GSTACK_STATE_ROOT\\nSLUG=$($B/gstack-slug --get SLUG 2>/dev/null); echo SLUG=$SLUG\\nCHECKPOINT_DIR=\\\"$GSTACK_STATE_ROOT/projects/$SLUG/checkpoints\\\"; ls -la \\\"$CHECKPOINT_DIR\\\"; find .gstack-home -type f\"},\"caller\":{\"type\":\"direct\"}}],\"container\":null,\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":6339,\"cache_read_input_tokens\":18298,\"cache_creation\":{\"ephemeral_5m_input_tokens\":6339,\"ephemeral_1h_input_tokens\":0},\"output_tokens\":8,\"service_tier\":\"standard\",\"inference_geo\":\"global\",\"speed\":\"standard\"},\"input_transformations\":[],\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"d1147327-49e3-4371-ab54-784d135573ac\",\"uuid\":\"c6cad35f-a6b6-4eb2-8741-39112ddc1cae\",\"timestamp\":\"2026-10-25T13:21:58.959Z\",\"request_id\":\"req_011CgP2Sw39kvpLEXQaPZoUk\",\"wire_tool_inputs\":{\"toolu_01EyWkH3v7Gu7cBDAgVzCiSu\":{\"command\":\"cd /tmp/gstack-paid-shard-O15pOD/tmp/skill-e2e-ctx-fragment-svqfPa; ls bin; export GSTACK_HOME=\\\"$PWD/.gstack-home\\\"; B=./bin\\nGSTACK_STATE_ROOT=$($B/gstack-paths --get GSTACK_STATE_ROOT); echo ROOT=$GSTACK_STATE_ROOT\\nSLUG=$($B/gstack-slug --get SLUG 2>/dev/null); echo SLUG=$SLUG\\nCHECKPOINT_DIR=\\\"$GSTACK_STATE_ROOT/projects/$SLUG/checkpoints\\\"; ls -la \\\"$CHECKPOINT_DIR\\\"; find .gstack-home -type f\"}}}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"tool_use_id\":\"toolu_01EyWkH3v7Gu7cBDAgVzCiSu\",\"type\":\"tool_result\",\"content\":\"gstack-config\\ngstack-learnings-log\\ngstack-learnings-search\\ngstack-paths\\ngstack-remote-identity.sh\\ngstack-repo-mode\\ngstack-slug\\ngstack-state-root.sh\\ngstack-timeline-log\\ngstack-timeline-read\\ngstack-update-check\\nROOT=/tmp/gstack-paid-shard-O15pOD/tmp/skill-e2e-ctx-fragment-svqfPa/.gstack-home\\nSLUG=skill-e2e-ctx-fragment-svqfPa\\ntotal 16\\ndrwxr-xr-x 2 user user 4096 Oct 25 13:21 .\\ndrwxr-xr-x 3 user user   60 Oct 25 13:21 ..\\n-rw-r--r-- 1 user user  157 Oct 25 13:21 20260101-120000-alpha-feature.md\\n-rw-r--r-- 1 user user  168 Oct 25 13:21 20260202-120000-middle-payments.md\\n-rw-r--r-- 1 user user  157 Oct 25 13:21 20260303-120000-omega-release.md\\n.gstack-home/projects/skill-e2e-ctx-fragment-svqfPa/checkpoints/20260101-120000-alpha-feature.md\\n.gstack-home/projects/skill-e2e-ctx-fragment-svqfPa/checkpoints/20260202-120000-middle-payments.md\\n.gstack-home/projects/skill-e2e-ctx-fragment-svqfPa/checkpoints/20260303-120000-omega-release.md\\n.gstack-home/slug-cache/_tmp_gstack-paid-shard-O15pOD_tmp_skill-e2e-ctx-fragment-svqfPa_.git\\n.gstack-home/slug-cache/_tmp_gstack-paid-shard-O15pOD_tmp_skill-e2e-ctx-fragment-svqfPa\",\"is_error\":false}]},\"parent_tool_use_id\":null,\"session_id\":\"d1147327-49e3-4371-ab54-784d135573ac\",\"uuid\":\"1beb6923-1013-4359-a964-2b06e7594c17\",\"timestamp\":\"2026-10-25T13:21:59.621Z\",\"tool_use_result\":{\"stdout\":\"gstack-config\\ngstack-learnings-log\\ngstack-learnings-search\\ngstack-paths\\ngstack-remote-identity.sh\\ngstack-repo-mode\\ngstack-slug\\ngstack-state-root.sh\\ngstack-timeline-log\\ngstack-timeline-read\\ngstack-update-check\\nROOT=/tmp/gstack-paid-shard-O15pOD/tmp/skill-e2e-ctx-fragment-svqfPa/.gstack-home\\nSLUG=skill-e2e-ctx-fragment-svqfPa\\ntotal 16\\ndrwxr-xr-x 2 user user 4096 Oct 25 13:21 .\\ndrwxr-xr-x 3 user user   60 Oct 25 13:21 ..\\n-rw-r--r-- 1 user user  157 Oct 25 13:21 20260101-120000-alpha-feature.md\\n-rw-r--r-- 1 user user  168 Oct 25 13:21 20260202-120000-middle-payments.md\\n-rw-r--r-- 1 user user  157 Oct 25 13:21 20260303-120000-omega-release.md\\n.gstack-home/projects/skill-e2e-ctx-fragment-svqfPa/checkpoints/20260101-120000-alpha-feature.md\\n.gstack-home/projects/skill-e2e-ctx-fragment-svqfPa/checkpoints/20260202-120000-middle-payments.md\\n.gstack-home/projects/skill-e2e-ctx-fragment-svqfPa/checkpoints/20260303-120000-omega-release.md\\n.gstack-home/slug-cache/_tmp_gstack-paid-shard-O15pOD_tmp_skill-e2e-ctx-fragment-svqfPa_.git\\n.gstack-home/slug-cache/_tmp_gstack-paid-shard-O15pOD_tmp_skill-e2e-ctx-fragment-svqfPa\",\"stderr\":\"\",\"interrupted\":false,\"isImage\":false,\"noOutputExpected\":false,\"bashEditDiff\":{\"files\":[{\"filePath\":\"/tmp/gstack-paid-shard-O15pOD/tmp/skill-e2e-ctx-fragment-svqfPa/.gstack-home/slug-cache/_tmp_gstack-paid-shard-O15pOD_tmp_skill-e2e-ctx-fragment-svqfPa\",\"hunks\":[{\"oldStart\":0,\"oldLines\":0,\"newStart\":1,\"newLines\":1,\"lines\":[\"+v2:skill-e2e-ctx-fragment-svqfPa\"]}],\"created\":true},{\"filePath\":\"/tmp/gstack-paid-shard-O15pOD/tmp/skill-e2e-ctx-fragment-svqfPa/.gstack-home/slug-cache/_tmp_gstack-paid-shard-O15pOD_tmp_skill-e2e-ctx-fragment-svqfPa_.git\",\"hunks\":[{\"oldStart\":0,\"oldLines\":0,\"newStart\":1,\"newLines\":1,\"lines\":[\"+v2:skill-e2e-ctx-fragment-svqfPa\"]}],\"created\":true}],\"moreFiles\":0,\"changedFiles\":[\"/tmp/gstack-paid-shard-O15pOD/tmp/skill-e2e-ctx-fragment-svqfPa/.gstack-home/slug-cache/_tmp_gstack-paid-shard-O15pOD_tmp_skill-e2e-ctx-fragment-svqfPa\",\"/tmp/gstack-paid-shard-O15pOD/tmp/skill-e2e-ctx-fragment-svqfPa/.gstack-home/slug-cache/_tmp_gstack-paid-shard-O15pOD_tmp_skill-e2e-ctx-fragment-svqfPa_.git\"]}}}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-5-5\",\"id\":\"msg_011CgP2TDR8mJC7cqYJgSDrA\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01E9t3wnBeL5brVygUoRfuoh\",\"name\":\"Bash\",\"input\":{\"command\":\"cd /tmp/gstack-paid-shard-O15pOD/tmp/skill-e2e-ctx-fragment-svqfPa/.gstack-home/projects/*/checkpoints; for f in *; do echo \\\"== $f\\\"; cat \\\"$f\\\"; done; cd ../../../..; export GSTACK_HOME=\\\"$PWD/.gstack-home\\\"; ls .gstack-home/projects/*/checkpoints/*.md | ./bin/gstack-slug --classify-checkpoints; echo rc=$?\"},\"caller\":{\"type\":\"direct\"}}],\"container\":null,\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":1039,\"cache_read_input_tokens\":24637,\"cache_creation\":{\"ephemeral_5m_input_tokens\":1039,\"ephemeral_1h_input_tokens\":0},\"output_tokens\":16,\"service_tier\":\"standard\",\"inference_geo\":\"global\",\"speed\":\"standard\"},\"input_transformations\":[],\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"d1147327-49e3-4371-ab54-784d135573ac\",\"uuid\":\"9fa3b490-1be5-4221-b712-c671e2422c46\",\"timestamp\":\"2026-10-25T13:22:01.775Z\",\"request_id\":\"req_011CgP2TD6nbxFnYubbkBVGv\",\"wire_tool_inputs\":{\"toolu_01E9t3wnBeL5brVygUoRfuoh\":{\"command\":\"cd /tmp/gstack-paid-shard-O15pOD/tmp/skill-e2e-ctx-fragment-svqfPa/.gstack-home/projects/*/checkpoints; for f in *; do echo \\\"== $f\\\"; cat \\\"$f\\\"; done; cd ../../../..; export GSTACK_HOME=\\\"$PWD/.gstack-home\\\"; ls .gstack-home/projects/*/checkpoints/*.md | ./bin/gstack-slug --classify-checkpoints; echo rc=$?\"}}}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"tool_use_id\":\"toolu_01E9t3wnBeL5brVygUoRfuoh\",\"type\":\"tool_result\",\"content\":\"== 20260101-120000-alpha-feature.md\\n---\\nstatus: in-progress\\nbranch: feat/alpha\\ntimestamp: 2026-01-01T12:00:00Z\\n---\\n## Working on: alpha feature\\n\\n### Summary\\nAlpha content FRAGMATCH_ALPHA_BUILD\\n== 20260202-120000-middle-payments.md\\n---\\nstatus: in-progress\\nbranch: feat/payments\\ntimestamp: 2026-02-02T12:00:00Z\\n---\\n## Working on: middle payments\\n\\n### Summary\\nPayments content FRAGMATCH_PAYMENTS_BUILD\\n== 20260303-120000-omega-release.md\\n---\\nstatus: in-progress\\nbranch: feat/omega\\ntimestamp: 2026-03-03T12:00:00Z\\n---\\n## Working on: omega release\\n\\n### Summary\\nOmega content FRAGMATCH_OMEGA_BUILD\\nunstamped\\t.gstack-home/projects/skill-e2e-ctx-fragment-svqfPa/checkpoints/20260101-120000-alpha-feature.md\\nunstamped\\t.gstack-home/projects/skill-e2e-ctx-fragment-svqfPa/checkpoints/20260202-120000-middle-payments.md\\nunstamped\\t.gstack-home/projects/skill-e2e-ctx-fragment-svqfPa/checkpoints/20260303-120000-omega-release.md\\nrc=0\",\"is_error\":false}]},\"parent_tool_use_id\":null,\"session_id\":\"d1147327-49e3-4371-ab54-784d135573ac\",\"uuid\":\"96853ca6-5db9-445b-ab54-ac9103ca336a\",\"timestamp\":\"2026-10-25T13:22:02.045Z\",\"tool_use_result\":{\"stdout\":\"== 20260101-120000-alpha-feature.md\\n---\\nstatus: in-progress\\nbranch: feat/alpha\\ntimestamp: 2026-01-01T12:00:00Z\\n---\\n## Working on: alpha feature\\n\\n### Summary\\nAlpha content FRAGMATCH_ALPHA_BUILD\\n== 20260202-120000-middle-payments.md\\n---\\nstatus: in-progress\\nbranch: feat/payments\\ntimestamp: 2026-02-02T12:00:00Z\\n---\\n## Working on: middle payments\\n\\n### Summary\\nPayments content FRAGMATCH_PAYMENTS_BUILD\\n== 20260303-120000-omega-release.md\\n---\\nstatus: in-progress\\nbranch: feat/omega\\ntimestamp: 2026-03-03T12:00:00Z\\n---\\n## Working on: omega release\\n\\n### Summary\\nOmega content FRAGMATCH_OMEGA_BUILD\\nunstamped\\t.gstack-home/projects/skill-e2e-ctx-fragment-svqfPa/checkpoints/20260101-120000-alpha-feature.md\\nunstamped\\t.gstack-home/projects/skill-e2e-ctx-fragment-svqfPa/checkpoints/20260202-120000-middle-payments.md\\nunstamped\\t.gstack-home/projects/skill-e2e-ctx-fragment-svqfPa/checkpoints/20260303-120000-omega-release.md\\nrc=0\",\"stderr\":\"\",\"interrupted\":false,\"isImage\":false,\"noOutputExpected\":false}}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-5-5\",\"id\":\"msg_011CgP2TPmyymz1BeaE3qngr\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"text\",\"text\":\"I loaded the checkpoint matching \\\"payments\\\": `20260202-120000-middle-payments.md`. It's the only one of the three saved checkpoints whose title contains that word. All three are unstamped and none come from another project, so there's no project mismatch.\\n\\n```\\nRESUMING CONTEXT\\n════════════════════════════════════════\\nTitle:       middle payments\\nBranch:      feat/payments\\nSaved:       Feb 2, 2026, 12:00 UTC\\nDuration:    (not recorded)\\nStatus:      in-progress\\n════════════════════════════════════════\\n\\n### Summary\\nPayments content FRAGMATCH_PAYMENTS_BUILD\\n\\n### Remaining Work\\n(none recorded in the checkpoint)\\n\\n### Notes\\n(none recorded)\\n```\\n\\n**Branch:** this context was saved on `feat/payments`, but you're on `main`. You may want to switch to `feat/payments` before continuing.\\n\\nThe other two checkpoints, which I did not load:\\n- `alpha feature` (`feat/alpha`, 2026-01-01)\\n- `omega release` (`feat/omega`, 2026-03-03), the newest\\n\\nYou asked me not to use AskUserQuestion, so I skipped the follow-up prompt. The checkpoint lists no remaining work. You can continue on the payments work, see the full saved file (it's only the short content above), or stop here.\"}],\"container\":null,\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":718,\"cache_read_input_tokens\":25676,\"cache_creation\":{\"ephemeral_5m_input_tokens\":718,\"ephemeral_1h_input_tokens\":0},\"output_tokens\":7,\"service_tier\":\"standard\",\"inference_geo\":\"global\",\"speed\":\"standard\"},\"input_transformations\":[],\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"d1147327-49e3-4371-ab54-784d135573ac\",\"uuid\":\"833cfc7c-5f23-4673-b45c-d7d70c0e3f35\",\"timestamp\":\"2026-10-25T13:22:06.611Z\",\"request_id\":\"req_011CgP2TPV8JNDtZyf4D3mYB\"}",
+    "{\"duration_api_ms\":11552,\"stop_reason\":\"end_turn\",\"session_id\":\"d1147327-49e3-4371-ab54-784d135573ac\",\"total_cost_usd\":0.1106026,\"usage\":{\"input_tokens\":8,\"cache_creation_input_tokens\":14352,\"cache_read_input_tokens\":80653,\"output_tokens\":1134,\"output_tokens_details\":{\"thinking_tokens\":86},\"server_tool_use\":{\"web_search_requests\":0,\"web_fetch_requests\":0},\"service_tier\":\"standard\",\"cache_creation\":{\"ephemeral_1h_input_tokens\":0,\"ephemeral_5m_input_tokens\":14352},\"inference_geo\":\"global\",\"iterations\":[{\"input_tokens\":2,\"output_tokens\":484,\"cache_read_input_tokens\":25676,\"cache_creation_input_tokens\":718,\"cache_creation\":{\"ephemeral_5m_input_tokens\":718,\"ephemeral_1h_input_tokens\":0},\"type\":\"message\"}],\"speed\":\"standard\",\"fallback_credit\":null},\"modelUsage\":{\"claude-opus-5-5\":{\"inputTokens\":8,\"outputTokens\":1134,\"cacheReadInputTokens\":80653,\"cacheCreationInputTokens\":14352,\"webSearchRequests\":0,\"costUSD\":0.1106026,\"contextWindow\":1000000,\"maxOutputTokens\":128000,\"thinkingTokens\":86,\"canonicalModel\":\"claude-opus-5-5\",\"provider\":\"firstParty\",\"costBasis\":\"list\"}},\"permission_denials\":[],\"terminal_reason\":\"completed\",\"fast_mode_state\":\"off\",\"fast_mode_disabled_reason\":\"sdk_opt_in_required\",\"subagent_stats\":{\"spawned\":0,\"requested\":{\"background\":0,\"foreground\":0,\"unset\":0},\"started_in_background\":0,\"max_depth\":0,\"spawned_by_subagents\":0,\"completed\":0,\"failed\":0,\"killed\":{\"parent\":0,\"user\":0,\"system\":0},\"refused\":{\"depth_limit\":0,\"concurrency_limit\":0,\"budget\":0},\"by_type\":{}},\"safety_stops\":0,\"is_error\":false,\"num_turns\":5,\"subtype\":\"success\",\"api_error_status\":null,\"result\":\"I loaded the checkpoint matching \\\"payments\\\": `20260202-120000-middle-payments.md`. It's the only one of the three saved checkpoints whose title contains that word. All three are unstamped and none come from another project, so there's no project mismatch.\\n\\n```\\nRESUMING CONTEXT\\n════════════════════════════════════════\\nTitle:       middle payments\\nBranch:      feat/payments\\nSaved:       Feb 2, 2026, 12:00 UTC\\nDuration:    (not recorded)\\nStatus:      in-progress\\n════════════════════════════════════════\\n\\n### Summary\\nPayments content FRAGMATCH_PAYMENTS_BUILD\\n\\n### Remaining Work\\n(none recorded in the checkpoint)\\n\\n### Notes\\n(none recorded)\\n```\\n\\n**Branch:** this context was saved on `feat/payments`, but you're on `main`. You may want to switch to `feat/payments` before continuing.\\n\\nThe other two checkpoints, which I did not load:\\n- `alpha feature` (`feat/alpha`, 2026-01-01)\\n- `omega release` (`feat/omega`, 2026-03-03), the newest\\n\\nYou asked me not to use AskUserQuestion, so I skipped the follow-up prompt. The checkpoint lists no remaining work. You can continue on the payments work, see the full saved file (it's only the short content above), or stop here.\",\"ttft_ms\":1839,\"type\":\"result\",\"duration_ms\":12657,\"uuid\":\"170f9374-7a93-44aa-9ce6-c54054faf5cc\",\"ttft_stream_ms\":902,\"time_to_request_ms\":199,\"first_content_frame_ms\":902,\"queued_turn_count\":0,\"result_index\":0}"
+   ]
+  },
+  {
+   "session_id": "8889e81b-d4bf-4eba-9ade-cd8805219972",
+   "ndjson": [
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-5-5\",\"id\":\"msg_011CgP2SpsKBS3fLGqXc3Bwm\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01DewVZTi4xMmGQYH14JLR52\",\"name\":\"Skill\",\"input\":{\"skill\":\"context-restore\",\"args\":\"payments\"},\"caller\":{\"type\":\"direct\"}}],\"container\":null,\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":6260,\"cache_read_input_tokens\":12042,\"cache_creation\":{\"ephemeral_5m_input_tokens\":6260,\"ephemeral_1h_input_tokens\":0},\"output_tokens\":16,\"service_tier\":\"standard\",\"inference_geo\":\"global\",\"speed\":\"standard\"},\"input_transformations\":[],\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"8889e81b-d4bf-4eba-9ade-cd8805219972\",\"uuid\":\"73df9278-656a-4350-9f2b-11f00fc83e23\",\"timestamp\":\"2026-10-25T13:21:55.982Z\",\"request_id\":\"req_011CgP2SpdBaxUoNcNgUqkdU\",\"wire_tool_inputs\":{\"toolu_01DewVZTi4xMmGQYH14JLR52\":{\"skill\":\"context-restore\",\"args\":\"payments\"}}}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"type\":\"tool_result\",\"tool_use_id\":\"toolu_01DewVZTi4xMmGQYH14JLR52\",\"content\":\"Launching skill: context-restore\"}]},\"parent_tool_use_id\":null,\"session_id\":\"8889e81b-d4bf-4eba-9ade-cd8805219972\",\"uuid\":\"d884aeed-d430-4d45-93b4-29855244b2ab\",\"timestamp\":\"2026-10-25T13:21:56.013Z\",\"tool_use_result\":{\"success\":true,\"commandName\":\"context-restore\",\"allowedTools\":[\"Bash\",\"Read\",\"Glob\",\"Grep\",\"AskUserQuestion\"]}}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-5-5\",\"id\":\"msg_011CgP2Swzw8ptggHdz4yMNm\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01YASRbf3r6DvzQdp2WBoi5v\",\"name\":\"Bash\",\"input\":{\"command\":\"cd /tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T; export GSTACK_HOME=\\\"$PWD/.gstack-home\\\"; ls bin; B=./bin\\nGSTACK_STATE_ROOT=$($B/gstack-paths --get GSTACK_STATE_ROOT); echo ROOT=$GSTACK_STATE_ROOT\\nSLUG=$($B/gstack-slug --get SLUG 2>/dev/null); echo SLUG=$SLUG\\nCHECKPOINT_DIR=\\\"$GSTACK_STATE_ROOT/projects/$SLUG/checkpoints\\\"; ls -la \\\"$CHECKPOINT_DIR\\\"; find \\\"$GSTACK_HOME\\\" -name \\\"*.md\\\"\",\"description\":\"Locate saved checkpoints\"},\"caller\":{\"type\":\"direct\"}}],\"container\":null,\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":6340,\"cache_read_input_tokens\":18302,\"cache_creation\":{\"ephemeral_5m_input_tokens\":6340,\"ephemeral_1h_input_tokens\":0},\"output_tokens\":8,\"service_tier\":\"standard\",\"inference_geo\":\"global\",\"speed\":\"standard\"},\"input_transformations\":[],\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"8889e81b-d4bf-4eba-9ade-cd8805219972\",\"uuid\":\"88f53b0b-92b0-433b-be2e-30a6e9c4f180\",\"timestamp\":\"2026-10-25T13:21:59.368Z\",\"request_id\":\"req_011CgP2SwkZ9Z8jUt6x2i38U\",\"wire_tool_inputs\":{\"toolu_01YASRbf3r6DvzQdp2WBoi5v\":{\"command\":\"cd /tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T; export GSTACK_HOME=\\\"$PWD/.gstack-home\\\"; ls bin; B=./bin\\nGSTACK_STATE_ROOT=$($B/gstack-paths --get GSTACK_STATE_ROOT); echo ROOT=$GSTACK_STATE_ROOT\\nSLUG=$($B/gstack-slug --get SLUG 2>/dev/null); echo SLUG=$SLUG\\nCHECKPOINT_DIR=\\\"$GSTACK_STATE_ROOT/projects/$SLUG/checkpoints\\\"; ls -la \\\"$CHECKPOINT_DIR\\\"; find \\\"$GSTACK_HOME\\\" -name \\\"*.md\\\"\",\"description\":\"Locate saved checkpoints\"}}}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"tool_use_id\":\"toolu_01YASRbf3r6DvzQdp2WBoi5v\",\"type\":\"tool_result\",\"content\":\"gstack-config\\ngstack-learnings-log\\ngstack-learnings-search\\ngstack-paths\\ngstack-remote-identity.sh\\ngstack-repo-mode\\ngstack-slug\\ngstack-state-root.sh\\ngstack-timeline-log\\ngstack-timeline-read\\ngstack-update-check\\nROOT=/tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home\\nSLUG=skill-e2e-ctx-fragment-mF7D6T\\ntotal 16\\ndrwxr-xr-x 2 user user 4096 Oct 25 13:21 .\\ndrwxr-xr-x 3 user user   60 Oct 25 13:21 ..\\n-rw-r--r-- 1 user user  157 Oct 25 13:21 20260101-120000-alpha-feature.md\\n-rw-r--r-- 1 user user  168 Oct 25 13:21 20260202-120000-middle-payments.md\\n-rw-r--r-- 1 user user  157 Oct 25 13:21 20260303-120000-omega-release.md\\n/tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home/projects/skill-e2e-ctx-fragment-mF7D6T/checkpoints/20260101-120000-alpha-feature.md\\n/tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home/projects/skill-e2e-ctx-fragment-mF7D6T/checkpoints/20260202-120000-middle-payments.md\\n/tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home/projects/skill-e2e-ctx-fragment-mF7D6T/checkpoints/20260303-120000-omega-release.md\",\"is_error\":false}]},\"parent_tool_use_id\":null,\"session_id\":\"8889e81b-d4bf-4eba-9ade-cd8805219972\",\"uuid\":\"d8dab3b1-3cb6-4750-a8b1-30e12536377f\",\"timestamp\":\"2026-10-25T13:21:59.932Z\",\"tool_use_result\":{\"stdout\":\"gstack-config\\ngstack-learnings-log\\ngstack-learnings-search\\ngstack-paths\\ngstack-remote-identity.sh\\ngstack-repo-mode\\ngstack-slug\\ngstack-state-root.sh\\ngstack-timeline-log\\ngstack-timeline-read\\ngstack-update-check\\nROOT=/tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home\\nSLUG=skill-e2e-ctx-fragment-mF7D6T\\ntotal 16\\ndrwxr-xr-x 2 user user 4096 Oct 25 13:21 .\\ndrwxr-xr-x 3 user user   60 Oct 25 13:21 ..\\n-rw-r--r-- 1 user user  157 Oct 25 13:21 20260101-120000-alpha-feature.md\\n-rw-r--r-- 1 user user  168 Oct 25 13:21 20260202-120000-middle-payments.md\\n-rw-r--r-- 1 user user  157 Oct 25 13:21 20260303-120000-omega-release.md\\n/tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home/projects/skill-e2e-ctx-fragment-mF7D6T/checkpoints/20260101-120000-alpha-feature.md\\n/tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home/projects/skill-e2e-ctx-fragment-mF7D6T/checkpoints/20260202-120000-middle-payments.md\\n/tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home/projects/skill-e2e-ctx-fragment-mF7D6T/checkpoints/20260303-120000-omega-release.md\",\"stderr\":\"\",\"interrupted\":false,\"isImage\":false,\"noOutputExpected\":false,\"bashEditDiff\":{\"files\":[{\"filePath\":\"/tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home/slug-cache/_tmp_gstack-paid-shard-XShYTB_tmp_skill-e2e-ctx-fragment-mF7D6T\",\"hunks\":[{\"oldStart\":0,\"oldLines\":0,\"newStart\":1,\"newLines\":1,\"lines\":[\"+v2:skill-e2e-ctx-fragment-mF7D6T\"]}],\"created\":true},{\"filePath\":\"/tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home/slug-cache/_tmp_gstack-paid-shard-XShYTB_tmp_skill-e2e-ctx-fragment-mF7D6T_.git\",\"hunks\":[{\"oldStart\":0,\"oldLines\":0,\"newStart\":1,\"newLines\":1,\"lines\":[\"+v2:skill-e2e-ctx-fragment-mF7D6T\"]}],\"created\":true}],\"moreFiles\":0,\"changedFiles\":[\"/tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home/slug-cache/_tmp_gstack-paid-shard-XShYTB_tmp_skill-e2e-ctx-fragment-mF7D6T\",\"/tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home/slug-cache/_tmp_gstack-paid-shard-XShYTB_tmp_skill-e2e-ctx-fragment-mF7D6T_.git\"]}}}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-5-5\",\"id\":\"msg_011CgP2TEimtnXFfvDRRUFde\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_0148N1vosxsxAic6XBPYrXGQ\",\"name\":\"Bash\",\"input\":{\"command\":\"cd /tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home/projects/*/checkpoints; for f in *; do echo \\\"== $f\\\"; cat \\\"$f\\\"; done\",\"description\":\"Read checkpoint files\"},\"caller\":{\"type\":\"direct\"}}],\"container\":null,\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":1098,\"cache_read_input_tokens\":24642,\"cache_creation\":{\"ephemeral_5m_input_tokens\":1098,\"ephemeral_1h_input_tokens\":0},\"output_tokens\":16,\"service_tier\":\"standard\",\"inference_geo\":\"global\",\"speed\":\"standard\"},\"input_transformations\":[],\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"8889e81b-d4bf-4eba-9ade-cd8805219972\",\"uuid\":\"83ba03a6-8c2c-4fb6-8a99-168b4da4b9ab\",\"timestamp\":\"2026-10-25T13:22:01.555Z\",\"request_id\":\"req_011CgP2TEStrRVjCJmLjGxUQ\",\"wire_tool_inputs\":{\"toolu_0148N1vosxsxAic6XBPYrXGQ\":{\"command\":\"cd /tmp/gstack-paid-shard-XShYTB/tmp/skill-e2e-ctx-fragment-mF7D6T/.gstack-home/projects/*/checkpoints; for f in *; do echo \\\"== $f\\\"; cat \\\"$f\\\"; done\",\"description\":\"Read checkpoint files\"}}}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"tool_use_id\":\"toolu_0148N1vosxsxAic6XBPYrXGQ\",\"type\":\"tool_result\",\"content\":\"== 20260101-120000-alpha-feature.md\\n---\\nstatus: in-progress\\nbranch: feat/alpha\\ntimestamp: 2026-01-01T12:00:00Z\\n---\\n## Working on: alpha feature\\n\\n### Summary\\nAlpha content FRAGMATCH_ALPHA_BUILD\\n== 20260202-120000-middle-payments.md\\n---\\nstatus: in-progress\\nbranch: feat/payments\\ntimestamp: 2026-02-02T12:00:00Z\\n---\\n## Working on: middle payments\\n\\n### Summary\\nPayments content FRAGMATCH_PAYMENTS_BUILD\\n== 20260303-120000-omega-release.md\\n---\\nstatus: in-progress\\nbranch: feat/omega\\ntimestamp: 2026-03-03T12:00:00Z\\n---\\n## Working on: omega release\\n\\n### Summary\\nOmega content FRAGMATCH_OMEGA_BUILD\",\"is_error\":false}]},\"parent_tool_use_id\":null,\"session_id\":\"8889e81b-d4bf-4eba-9ade-cd8805219972\",\"uuid\":\"686ae24d-a3eb-4707-ae3b-330574244652\",\"timestamp\":\"2026-10-25T13:22:01.697Z\",\"tool_use_result\":{\"stdout\":\"== 20260101-120000-alpha-feature.md\\n---\\nstatus: in-progress\\nbranch: feat/alpha\\ntimestamp: 2026-01-01T12:00:00Z\\n---\\n## Working on: alpha feature\\n\\n### Summary\\nAlpha content FRAGMATCH_ALPHA_BUILD\\n== 20260202-120000-middle-payments.md\\n---\\nstatus: in-progress\\nbranch: feat/payments\\ntimestamp: 2026-02-02T12:00:00Z\\n---\\n## Working on: middle payments\\n\\n### Summary\\nPayments content FRAGMATCH_PAYMENTS_BUILD\\n== 20260303-120000-omega-release.md\\n---\\nstatus: in-progress\\nbranch: feat/omega\\ntimestamp: 2026-03-03T12:00:00Z\\n---\\n## Working on: omega release\\n\\n### Summary\\nOmega content FRAGMATCH_OMEGA_BUILD\",\"stderr\":\"\",\"interrupted\":false,\"isImage\":false,\"noOutputExpected\":false}}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-5-5\",\"id\":\"msg_011CgP2TNGCLmG6DD1SMJXr4\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"text\",\"text\":\"I loaded the checkpoint matching \\\"payments\\\": `20260202-120000-middle-payments.md`. Three checkpoints were saved in this project (alpha-feature, middle-payments, omega-release), and only this one has \\\"payments\\\" in its title. Omega-release is newer, but you asked for this one by name.\\n\\n```\\nRESUMING CONTEXT\\n════════════════════════════════════════\\nTitle:       middle payments\\nBranch:      feat/payments\\nSaved:       2026-02-02 12:00:00 UTC\\nDuration:    (not recorded)\\nStatus:      in-progress\\n════════════════════════════════════════\\n\\n### Summary\\nPayments content FRAGMATCH_PAYMENTS_BUILD\\n\\n### Remaining Work\\n(none recorded)\\n\\n### Notes\\n(none recorded)\\n```\\n\\n- **Branch:** this context was saved on `feat/payments`, but you're on `main`. You may want to switch branches before continuing.\\n- **Project check:** the checkpoint has no project identity stamp. That's fine here, because no other project has saved checkpoints in this folder.\\n\\nI didn't change any code, and I skipped the usual follow-up question as you asked. Your options are to keep working on this context, see the full file (it's short and shown above in full), or stop here.\"}],\"container\":null,\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":635,\"cache_read_input_tokens\":25740,\"cache_creation\":{\"ephemeral_5m_input_tokens\":635,\"ephemeral_1h_input_tokens\":0},\"output_tokens\":2,\"service_tier\":\"standard\",\"inference_geo\":\"global\",\"speed\":\"standard\"},\"input_transformations\":[],\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"8889e81b-d4bf-4eba-9ade-cd8805219972\",\"uuid\":\"70979650-0199-44e3-956e-0824d1de507a\",\"timestamp\":\"2026-10-25T13:22:06.359Z\",\"request_id\":\"req_011CgP2TN24xUJz1E2rmAPb1\"}",
+    "{\"duration_api_ms\":11304,\"stop_reason\":\"end_turn\",\"session_id\":\"8889e81b-d4bf-4eba-9ade-cd8805219972\",\"total_cost_usd\":0.10918220000000001,\"usage\":{\"input_tokens\":8,\"cache_creation_input_tokens\":14333,\"cache_read_input_tokens\":80726,\"output_tokens\":1067,\"output_tokens_details\":{\"thinking_tokens\":71},\"server_tool_use\":{\"web_search_requests\":0,\"web_fetch_requests\":0},\"service_tier\":\"standard\",\"cache_creation\":{\"ephemeral_1h_input_tokens\":0,\"ephemeral_5m_input_tokens\":14333},\"inference_geo\":\"global\",\"iterations\":[{\"input_tokens\":2,\"output_tokens\":439,\"cache_read_input_tokens\":25740,\"cache_creation_input_tokens\":635,\"cache_creation\":{\"ephemeral_5m_input_tokens\":635,\"ephemeral_1h_input_tokens\":0},\"type\":\"message\"}],\"speed\":\"standard\",\"fallback_credit\":null},\"modelUsage\":{\"claude-opus-5-5\":{\"inputTokens\":8,\"outputTokens\":1067,\"cacheReadInputTokens\":80726,\"cacheCreationInputTokens\":14333,\"webSearchRequests\":0,\"costUSD\":0.10918220000000001,\"contextWindow\":1000000,\"maxOutputTokens\":128000,\"thinkingTokens\":71,\"canonicalModel\":\"claude-opus-5-5\",\"provider\":\"firstParty\",\"costBasis\":\"list\"}},\"permission_denials\":[],\"terminal_reason\":\"completed\",\"fast_mode_state\":\"off\",\"fast_mode_disabled_reason\":\"sdk_opt_in_required\",\"subagent_stats\":{\"spawned\":0,\"requested\":{\"background\":0,\"foreground\":0,\"unset\":0},\"started_in_background\":0,\"max_depth\":0,\"spawned_by_subagents\":0,\"completed\":0,\"failed\":0,\"killed\":{\"parent\":0,\"user\":0,\"system\":0},\"refused\":{\"depth_limit\":0,\"concurrency_limit\":0,\"budget\":0},\"by_type\":{}},\"safety_stops\":0,\"is_error\":false,\"num_turns\":5,\"subtype\":\"success\",\"api_error_status\":null,\"result\":\"I loaded the checkpoint matching \\\"payments\\\": `20260202-120000-middle-payments.md`. Three checkpoints were saved in this project (alpha-feature, middle-payments, omega-release), and only this one has \\\"payments\\\" in its title. Omega-release is newer, but you asked for this one by name.\\n\\n```\\nRESUMING CONTEXT\\n════════════════════════════════════════\\nTitle:       middle payments\\nBranch:      feat/payments\\nSaved:       2026-02-02 12:00:00 UTC\\nDuration:    (not recorded)\\nStatus:      in-progress\\n════════════════════════════════════════\\n\\n### Summary\\nPayments content FRAGMATCH_PAYMENTS_BUILD\\n\\n### Remaining Work\\n(none recorded)\\n\\n### Notes\\n(none recorded)\\n```\\n\\n- **Branch:** this context was saved on `feat/payments`, but you're on `main`. You may want to switch branches before continuing.\\n- **Project check:** the checkpoint has no project identity stamp. That's fine here, because no other project has saved checkpoints in this folder.\\n\\nI didn't change any code, and I skipped the usual follow-up question as you asked. Your options are to keep working on this context, see the full file (it's short and shown above in full), or stop here.\",\"ttft_ms\":1797,\"type\":\"result\",\"duration_ms\":12189,\"uuid\":\"3003a3de-20c9-4a22-8387-f4c23597d9c3\",\"ttft_stream_ms\":873,\"time_to_request_ms\":179,\"first_content_frame_ms\":874,\"queued_turn_count\":0,\"result_index\":0}"
+   ]
+  },
+  {
+   "session_id": "6d684089-4cb5-4016-bc86-203a9a8cdab3",
+   "ndjson": [
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-5-5\",\"id\":\"msg_011CgP2SoyF1qZULHrLjmM3K\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01X9nJt9XVFgAKfbeHTY7jhD\",\"name\":\"Skill\",\"input\":{\"skill\":\"context-restore\",\"args\":\"payments\"},\"caller\":{\"type\":\"direct\"}}],\"container\":null,\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":6260,\"cache_read_input_tokens\":12042,\"cache_creation\":{\"ephemeral_5m_input_tokens\":6260,\"ephemeral_1h_input_tokens\":0},\"output_tokens\":16,\"service_tier\":\"standard\",\"inference_geo\":\"global\",\"speed\":\"standard\"},\"input_transformations\":[],\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"6d684089-4cb5-4016-bc86-203a9a8cdab3\",\"uuid\":\"7756244d-e82e-4aeb-9437-732412ccb96f\",\"timestamp\":\"2026-10-25T13:21:56.218Z\",\"request_id\":\"req_011CgP2Soi7ME2b1kXN8NCRG\",\"wire_tool_inputs\":{\"toolu_01X9nJt9XVFgAKfbeHTY7jhD\":{\"skill\":\"context-restore\",\"args\":\"payments\"}}}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"type\":\"tool_result\",\"tool_use_id\":\"toolu_01X9nJt9XVFgAKfbeHTY7jhD\",\"content\":\"Launching skill: context-restore\"}]},\"parent_tool_use_id\":null,\"session_id\":\"6d684089-4cb5-4016-bc86-203a9a8cdab3\",\"uuid\":\"376841c6-0fe2-417f-8da9-5d79c52f0694\",\"timestamp\":\"2026-10-25T13:21:56.247Z\",\"tool_use_result\":{\"success\":true,\"commandName\":\"context-restore\",\"allowedTools\":[\"Bash\",\"Read\",\"Glob\",\"Grep\",\"AskUserQuestion\"]}}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-5-5\",\"id\":\"msg_011CgP2Sy7Apg39RghKe7nhs\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_016oZY36mo3ibs1cL3rNt82T\",\"name\":\"Bash\",\"input\":{\"command\":\"cd /tmp/gstack-paid-shard-V2G2le/tmp/skill-e2e-ctx-fragment-GZoJH5; ls bin; export GSTACK_HOME=\\\"$PWD/.gstack-home\\\"; B=./bin\\nGSTACK_STATE_ROOT=$($B/gstack-paths --get GSTACK_STATE_ROOT 2>&1); echo \\\"ROOT=$GSTACK_STATE_ROOT\\\"\\nSLUG=$($B/gstack-slug --get SLUG 2>&1); echo \\\"SLUG=$SLUG\\\"\\nfind .gstack-home -name \\\"*.md\\\" | sort -r\",\"description\":\"Locate gstack state and checkpoints\"},\"caller\":{\"type\":\"direct\"}}],\"container\":null,\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":6340,\"cache_read_input_tokens\":18302,\"cache_creation\":{\"ephemeral_5m_input_tokens\":6340,\"ephemeral_1h_input_tokens\":0},\"output_tokens\":8,\"service_tier\":\"standard\",\"inference_geo\":\"global\",\"speed\":\"standard\"},\"input_transformations\":[],\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"6d684089-4cb5-4016-bc86-203a9a8cdab3\",\"uuid\":\"46203839-ca20-4fe1-a98b-d965e965f665\",\"timestamp\":\"2026-10-25T13:22:00.005Z\",\"request_id\":\"req_011CgP2SxnaBEsoXQB2CY6kN\",\"wire_tool_inputs\":{\"toolu_016oZY36mo3ibs1cL3rNt82T\":{\"command\":\"cd /tmp/gstack-paid-shard-V2G2le/tmp/skill-e2e-ctx-fragment-GZoJH5; ls bin; export GSTACK_HOME=\\\"$PWD/.gstack-home\\\"; B=./bin\\nGSTACK_STATE_ROOT=$($B/gstack-paths --get GSTACK_STATE_ROOT 2>&1); echo \\\"ROOT=$GSTACK_STATE_ROOT\\\"\\nSLUG=$($B/gstack-slug --get SLUG 2>&1); echo \\\"SLUG=$SLUG\\\"\\nfind .gstack-home -name \\\"*.md\\\" | sort -r\",\"description\":\"Locate gstack state and checkpoints\"}}}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"tool_use_id\":\"toolu_016oZY36mo3ibs1cL3rNt82T\",\"type\":\"tool_result\",\"content\":\"gstack-config\\ngstack-learnings-log\\ngstack-learnings-search\\ngstack-paths\\ngstack-remote-identity.sh\\ngstack-repo-mode\\ngstack-slug\\ngstack-state-root.sh\\ngstack-timeline-log\\ngstack-timeline-read\\ngstack-update-check\\nROOT=/tmp/gstack-paid-shard-V2G2le/tmp/skill-e2e-ctx-fragment-GZoJH5/.gstack-home\\nSLUG=skill-e2e-ctx-fragment-GZoJH5\\n.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260303-120000-omega-release.md\\n.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260202-120000-middle-payments.md\\n.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260101-120000-alpha-feature.md\",\"is_error\":false}]},\"parent_tool_use_id\":null,\"session_id\":\"6d684089-4cb5-4016-bc86-203a9a8cdab3\",\"uuid\":\"d7748b04-b48b-498b-89e5-ce4df4e6d58c\",\"timestamp\":\"2026-10-25T13:22:00.921Z\",\"tool_use_result\":{\"stdout\":\"gstack-config\\ngstack-learnings-log\\ngstack-learnings-search\\ngstack-paths\\ngstack-remote-identity.sh\\ngstack-repo-mode\\ngstack-slug\\ngstack-state-root.sh\\ngstack-timeline-log\\ngstack-timeline-read\\ngstack-update-check\\nROOT=/tmp/gstack-paid-shard-V2G2le/tmp/skill-e2e-ctx-fragment-GZoJH5/.gstack-home\\nSLUG=skill-e2e-ctx-fragment-GZoJH5\\n.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260303-120000-omega-release.md\\n.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260202-120000-middle-payments.md\\n.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260101-120000-alpha-feature.md\",\"stderr\":\"\",\"interrupted\":false,\"isImage\":false,\"noOutputExpected\":false,\"bashEditDiff\":{\"files\":[{\"filePath\":\"/tmp/gstack-paid-shard-V2G2le/tmp/skill-e2e-ctx-fragment-GZoJH5/.gstack-home/slug-cache/_tmp_gstack-paid-shard-V2G2le_tmp_skill-e2e-ctx-fragment-GZoJH5\",\"hunks\":[{\"oldStart\":0,\"oldLines\":0,\"newStart\":1,\"newLines\":1,\"lines\":[\"+v2:skill-e2e-ctx-fragment-GZoJH5\"]}],\"created\":true},{\"filePath\":\"/tmp/gstack-paid-shard-V2G2le/tmp/skill-e2e-ctx-fragment-GZoJH5/.gstack-home/slug-cache/_tmp_gstack-paid-shard-V2G2le_tmp_skill-e2e-ctx-fragment-GZoJH5_.git\",\"hunks\":[{\"oldStart\":0,\"oldLines\":0,\"newStart\":1,\"newLines\":1,\"lines\":[\"+v2:skill-e2e-ctx-fragment-GZoJH5\"]}],\"created\":true}],\"moreFiles\":0,\"changedFiles\":[\"/tmp/gstack-paid-shard-V2G2le/tmp/skill-e2e-ctx-fragment-GZoJH5/.gstack-home/slug-cache/_tmp_gstack-paid-shard-V2G2le_tmp_skill-e2e-ctx-fragment-GZoJH5\",\"/tmp/gstack-paid-shard-V2G2le/tmp/skill-e2e-ctx-fragment-GZoJH5/.gstack-home/slug-cache/_tmp_gstack-paid-shard-V2G2le_tmp_skill-e2e-ctx-fragment-GZoJH5_.git\"]}}}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-5-5\",\"id\":\"msg_011CgP2TJxWctw7y2Vz5qBJH\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"tool_use\",\"id\":\"toolu_01FMDVzkCc6y3vknp52LsMBF\",\"name\":\"Bash\",\"input\":{\"command\":\"cd /tmp/gstack-paid-shard-V2G2le/tmp/skill-e2e-ctx-fragment-GZoJH5; export GSTACK_HOME=\\\"$PWD/.gstack-home\\\"; D=.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints\\nls $D/*.md | ./bin/gstack-slug --classify-checkpoints 2>&1; echo ---; cat $D/20260202-120000-middle-payments.md; echo ---; head -12 $D/20260303-120000-omega-release.md $D/20260101-120000-alpha-feature.md\",\"description\":\"Classify checkpoints and read payments file\"},\"caller\":{\"type\":\"direct\"}}],\"container\":null,\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":718,\"cache_read_input_tokens\":24642,\"cache_creation\":{\"ephemeral_5m_input_tokens\":718,\"ephemeral_1h_input_tokens\":0},\"output_tokens\":16,\"service_tier\":\"standard\",\"inference_geo\":\"global\",\"speed\":\"standard\"},\"input_transformations\":[],\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"6d684089-4cb5-4016-bc86-203a9a8cdab3\",\"uuid\":\"bf17bfed-5ea2-42c9-8b0d-df4adf7e20ff\",\"timestamp\":\"2026-10-25T13:22:03.980Z\",\"request_id\":\"req_011CgP2TJg9vKd8HjaMC2abX\",\"wire_tool_inputs\":{\"toolu_01FMDVzkCc6y3vknp52LsMBF\":{\"command\":\"cd /tmp/gstack-paid-shard-V2G2le/tmp/skill-e2e-ctx-fragment-GZoJH5; export GSTACK_HOME=\\\"$PWD/.gstack-home\\\"; D=.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints\\nls $D/*.md | ./bin/gstack-slug --classify-checkpoints 2>&1; echo ---; cat $D/20260202-120000-middle-payments.md; echo ---; head -12 $D/20260303-120000-omega-release.md $D/20260101-120000-alpha-feature.md\",\"description\":\"Classify checkpoints and read payments file\"}}}",
+    "{\"type\":\"user\",\"message\":{\"role\":\"user\",\"content\":[{\"tool_use_id\":\"toolu_01FMDVzkCc6y3vknp52LsMBF\",\"type\":\"tool_result\",\"content\":\"unstamped\\t.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260101-120000-alpha-feature.md\\nunstamped\\t.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260202-120000-middle-payments.md\\nunstamped\\t.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260303-120000-omega-release.md\\n---\\n---\\nstatus: in-progress\\nbranch: feat/payments\\ntimestamp: 2026-02-02T12:00:00Z\\n---\\n## Working on: middle payments\\n\\n### Summary\\nPayments content FRAGMATCH_PAYMENTS_BUILD\\n---\\n==> .gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260303-120000-omega-release.md <==\\n---\\nstatus: in-progress\\nbranch: feat/omega\\ntimestamp: 2026-03-03T12:00:00Z\\n---\\n## Working on: omega release\\n\\n### Summary\\nOmega content FRAGMATCH_OMEGA_BUILD\\n\\n==> .gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260101-120000-alpha-feature.md <==\\n---\\nstatus: in-progress\\nbranch: feat/alpha\\ntimestamp: 2026-01-01T12:00:00Z\\n---\\n## Working on: alpha feature\\n\\n### Summary\\nAlpha content FRAGMATCH_ALPHA_BUILD\",\"is_error\":false}]},\"parent_tool_use_id\":null,\"session_id\":\"6d684089-4cb5-4016-bc86-203a9a8cdab3\",\"uuid\":\"21ec78ab-abbc-4ea3-8ba6-544d6b05fc91\",\"timestamp\":\"2026-10-25T13:22:04.223Z\",\"tool_use_result\":{\"stdout\":\"unstamped\\t.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260101-120000-alpha-feature.md\\nunstamped\\t.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260202-120000-middle-payments.md\\nunstamped\\t.gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260303-120000-omega-release.md\\n---\\n---\\nstatus: in-progress\\nbranch: feat/payments\\ntimestamp: 2026-02-02T12:00:00Z\\n---\\n## Working on: middle payments\\n\\n### Summary\\nPayments content FRAGMATCH_PAYMENTS_BUILD\\n---\\n==> .gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260303-120000-omega-release.md <==\\n---\\nstatus: in-progress\\nbranch: feat/omega\\ntimestamp: 2026-03-03T12:00:00Z\\n---\\n## Working on: omega release\\n\\n### Summary\\nOmega content FRAGMATCH_OMEGA_BUILD\\n\\n==> .gstack-home/projects/skill-e2e-ctx-fragment-GZoJH5/checkpoints/20260101-120000-alpha-feature.md <==\\n---\\nstatus: in-progress\\nbranch: feat/alpha\\ntimestamp: 2026-01-01T12:00:00Z\\n---\\n## Working on: alpha feature\\n\\n### Summary\\nAlpha content FRAGMATCH_ALPHA_BUILD\",\"stderr\":\"\",\"interrupted\":false,\"isImage\":false,\"noOutputExpected\":false}}",
+    "{\"type\":\"assistant\",\"message\":{\"model\":\"claude-opus-5-5\",\"id\":\"msg_011CgP2TZ93aTmrLHgHkg2Je\",\"type\":\"message\",\"role\":\"assistant\",\"content\":[{\"type\":\"text\",\"text\":\"I loaded the checkpoint `20260202-120000-middle-payments.md`. It's the only one of the three saved checkpoints whose title contains \\\"payments\\\". It isn't the newest: `omega-release` (2026-03-03) is, so I chose this one because its title matched, not because it was the latest.\\n\\n```\\nRESUMING CONTEXT\\n════════════════════════════════════════\\nTitle:       middle payments\\nBranch:      feat/payments\\nSaved:       Feb 2, 2026, 12:00 UTC\\nDuration:    not recorded\\nStatus:      in-progress\\n════════════════════════════════════════\\n\\n### Summary\\nPayments content FRAGMATCH_PAYMENTS_BUILD\\n\\n### Remaining Work\\nNone recorded in the checkpoint.\\n\\n### Notes\\nNone recorded.\\n```\\n\\n- **Branch:** this was saved on `feat/payments`, but you're on `main`. You may want to switch branches before continuing.\\n- **Project check:** none of the checkpoints have a project identity stamp, but none belong to another project, so this one counts as this project's.\\n- **Other checkpoints:** `alpha feature` on `feat/alpha` (2026-01-01) and `omega release` on `feat/omega` (2026-03-03).\\n\\nYou asked me not to use AskUserQuestion, so I skipped the skill's usual \\\"what next?\\\" prompt. I made no code changes; restoring only reads the saved files.\"}],\"container\":null,\"stop_reason\":null,\"stop_sequence\":null,\"stop_details\":null,\"usage\":{\"input_tokens\":2,\"cache_creation_input_tokens\":859,\"cache_read_input_tokens\":25360,\"cache_creation\":{\"ephemeral_5m_input_tokens\":859,\"ephemeral_1h_input_tokens\":0},\"output_tokens\":8,\"service_tier\":\"standard\",\"inference_geo\":\"global\",\"speed\":\"standard\"},\"input_transformations\":[],\"diagnostics\":null,\"context_management\":null},\"parent_tool_use_id\":null,\"session_id\":\"6d684089-4cb5-4016-bc86-203a9a8cdab3\",\"uuid\":\"e8864f23-50e9-4a59-ab4a-b984590d2559\",\"timestamp\":\"2026-10-25T13:22:09.545Z\",\"request_id\":\"req_011CgP2TYqh2XjMw3Nc9cuQu\"}",
+    "{\"duration_api_ms\":14272,\"stop_reason\":\"end_turn\",\"session_id\":\"6d684089-4cb5-4016-bc86-203a9a8cdab3\",\"total_cost_usd\":0.1112262,\"usage\":{\"input_tokens\":8,\"cache_creation_input_tokens\":14177,\"cache_read_input_tokens\":80346,\"output_tokens\":1212,\"output_tokens_details\":{\"thinking_tokens\":88},\"server_tool_use\":{\"web_search_requests\":0,\"web_fetch_requests\":0},\"service_tier\":\"standard\",\"cache_creation\":{\"ephemeral_1h_input_tokens\":0,\"ephemeral_5m_input_tokens\":14177},\"inference_geo\":\"global\",\"iterations\":[{\"input_tokens\":2,\"output_tokens\":501,\"cache_read_input_tokens\":25360,\"cache_creation_input_tokens\":859,\"cache_creation\":{\"ephemeral_5m_input_tokens\":859,\"ephemeral_1h_input_tokens\":0},\"type\":\"message\"}],\"speed\":\"standard\",\"fallback_credit\":null},\"modelUsage\":{\"claude-opus-5-5\":{\"inputTokens\":8,\"outputTokens\":1212,\"cacheReadInputTokens\":80346,\"cacheCreationInputTokens\":14177,\"webSearchRequests\":0,\"costUSD\":0.1112262,\"contextWindow\":1000000,\"maxOutputTokens\":128000,\"thinkingTokens\":88,\"canonicalModel\":\"claude-opus-5-5\",\"provider\":\"firstParty\",\"costBasis\":\"list\"}},\"permission_denials\":[],\"terminal_reason\":\"completed\",\"fast_mode_state\":\"off\",\"fast_mode_disabled_reason\":\"sdk_opt_in_required\",\"subagent_stats\":{\"spawned\":0,\"requested\":{\"background\":0,\"foreground\":0,\"unset\":0},\"started_in_background\":0,\"max_depth\":0,\"spawned_by_subagents\":0,\"completed\":0,\"failed\":0,\"killed\":{\"parent\":0,\"user\":0,\"system\":0},\"refused\":{\"depth_limit\":0,\"concurrency_limit\":0,\"budget\":0},\"by_type\":{}},\"safety_stops\":0,\"is_error\":false,\"num_turns\":5,\"subtype\":\"success\",\"api_error_status\":null,\"result\":\"I loaded the checkpoint `20260202-120000-middle-payments.md`. It's the only one of the three saved checkpoints whose title contains \\\"payments\\\". It isn't the newest: `omega-release` (2026-03-03) is, so I chose this one because its title matched, not because it was the latest.\\n\\n```\\nRESUMING CONTEXT\\n════════════════════════════════════════\\nTitle:       middle payments\\nBranch:      feat/payments\\nSaved:       Feb 2, 2026, 12:00 UTC\\nDuration:    not recorded\\nStatus:      in-progress\\n════════════════════════════════════════\\n\\n### Summary\\nPayments content FRAGMATCH_PAYMENTS_BUILD\\n\\n### Remaining Work\\nNone recorded in the checkpoint.\\n\\n### Notes\\nNone recorded.\\n```\\n\\n- **Branch:** this was saved on `feat/payments`, but you're on `main`. You may want to switch branches before continuing.\\n- **Project check:** none of the checkpoints have a project identity stamp, but none belong to another project, so this one counts as this project's.\\n- **Other checkpoints:** `alpha feature` on `feat/alpha` (2026-01-01) and `omega release` on `feat/omega` (2026-03-03).\\n\\nYou asked me not to use AskUserQuestion, so I skipped the skill's usual \\\"what next?\\\" prompt. I made no code changes; restoring only reads the saved files.\",\"ttft_ms\":2268,\"type\":\"result\",\"duration_ms\":15609,\"uuid\":\"39ae4b99-23c4-4444-b94e-4cf48961a993\",\"ttft_stream_ms\":1914,\"time_to_request_ms\":192,\"first_content_frame_ms\":1915,\"queued_turn_count\":0,\"result_index\":0}"
+   ]
+  }
+ ]
+}

test/gstack-ff-update.test.ts

@@ -336,10 +336,12 @@ describe.skipIf(process.platform === 'win32')('gstack-ff-update: the install-sco
   test('the heartbeat and its sleeper are reaped when the bin exits', () => {
     const fx = makeFixture();
     release(fx, { 'notes.txt': 'top UPSTREAM\nmiddle\nbottom\n' });
-    const before = execFileSync('bash', ['-c', 'pgrep -f "^sleep 300$" | wc -l'], { encoding: 'utf8', timeout: 30_000 }).trim();
-    expect(ff(fx).status).toBe(0);
-    const after = execFileSync('bash', ['-c', 'sleep 0.3; pgrep -f "^sleep 300$" | wc -l'], { encoding: 'utf8', timeout: 30_000 }).trim();
-    expect(after).toBe(before);
+    const beat = String(3000 + (process.pid % 997));
+    const sleepers = (delay = '') => execFileSync('bash', ['-c', `${delay}pgrep -f "^sleep ${beat}$" | wc -l`], { encoding: 'utf8', timeout: 30_000 }).trim();
+    const before = sleepers();
+    expect(before).toBe('0');
+    expect(ff(fx, 'origin/main', { GSTACK_FF_HEARTBEAT_SECONDS: beat }).status).toBe(0);
+    expect(sleepers('sleep 0.3; ')).toBe(before);
   });
 });
 

test/gstack-skill-start.test.ts

@@ -932,7 +932,7 @@ describe('gstack-skill-start install admission', () => {
     const elsewhere = verdict(gh, {}, path.join(gh, 'named.json'));
     fs.rmSync(path.join(gh, 'install-verdicts'), { recursive: true, force: true });
     expect(start(gh, 'autoplan', { ...capy, GSTACK_INSTALL_VERDICT: elsewhere }).status).toBe(0);
-  });
+  }, 20_000);
 
   test('an installed runtime root follows its .source-path to the source checkout for the verdict and HEAD', () => {
     const gh = scratch();

test/helpers/carve-guards.ts

@@ -736,8 +736,8 @@ do not launch the downstream skill or open a browser.`,
   // (D3A: read-on-demand doctrine, requiredReads-guarded + loading eval)
   'design-html': {
     skill: 'design-html',
-    expectedSections: ['doctrine.md', 'pretext-patterns.md', 'detector-install-offer.md'],
-    requiredReads: ['doctrine.md', 'pretext-patterns.md'],
+    expectedSections: ['doctrine.md', 'pretext-patterns.md', 'craft.md', 'detector-install-offer.md'],
+    requiredReads: ['doctrine.md', 'pretext-patterns.md', 'craft.md'],
     scenario:
       'Walk /design-html in SIMULATION — do not run bash, start servers, launch a browser, or take screenshots. Treat Step 0 as already resolved: no CEO plan, no approved mockup, no variants, no DESIGN.md, no prior finalized.html — freeform mode (Case C option D), screen name "pricing", the user wants a pricing page for a developer-tools SaaS (dark, dense, three tiers, monospace-leaning). Do NOT use AskUserQuestion — proceed with the stated assumptions. Read each pointed section before doing its step, then execute Steps 1-3: produce the implementation spec, state the chosen Pretext tier and why, and generate the complete Pretext-native HTML — include the HTML in your report instead of writing files. Stop there: skip Step 3.5, Step 4, and Step 5.',
     staticInvariants: {
@@ -759,7 +759,7 @@ do not launch the downstream skill or open a browser.`,
       gateAfterStop: undefined, // operational skill, no plan-mode gate
     },
     behavioral: 'prompt',
-    maxSkeletonBytes: 56_200, // measured 55,262 (2026-09-09): the detector install offer pointer + its sections-table row (the brief itself lives in sections/detector-install-offer.md); before that 54,545 for the review-cycle trust prose, the Slop Gate's Decisions-Log clause, and the blacklist header's override sentence + Honest Work wave merged (W2/W3 gate integrity in review/ship, W4 kernel + DONE line in every tier-2+ skill, W5/W7 in spec/ship): re-measured on the integrated tree (2026-10-10).
+    maxSkeletonBytes: 56_200, // measured 55,786 (2026-10-25): the craft section pointer, its index row and the Step 4 craft-check line (GSTA-92); before that 55,262 (2026-09-09): the detector install offer pointer + its sections-table row (the brief itself lives in sections/detector-install-offer.md); before that 54,545 for the review-cycle trust prose, the Slop Gate's Decisions-Log clause, and the blacklist header's override sentence + Honest Work wave merged (W2/W3 gate integrity in review/ship, W4 kernel + DONE line in every tier-2+ skill, W5/W7 in spec/ship): re-measured on the integrated tree (2026-10-10).
     minUnionBytes: 57_500, // Phase 4 wave 4; measured union 58,682
     mustContain: ["Don't make me think", "Users scan, they don't read", 'The Goodwill Reservoir', 'PRETEXT API CHEATSHEET', 'Pattern 3: Text around obstacles'],
   },

test/helpers/context-restore-fragment.ts

@@ -0,0 +1,34 @@
+/**
+ * Verdict for the context-restore-fragment-match E2E: which checkpoint did the
+ * agent restore? The evidence is the agent's own report (top-level assistant
+ * text plus the final result), never tool output. Restore has to inspect the
+ * candidates to find the one whose title matches, so every checkpoint body can
+ * appear in Bash or Read output without being the restored context.
+ */
+export interface FragmentRestoreRun {
+  output?: string;
+  transcript?: any[];
+}
+
+export function agentReport(result: FragmentRestoreRun): string {
+  const parts: string[] = [];
+  for (const event of result.transcript || []) {
+    if (event?.type !== 'assistant' || event.parent_tool_use_id) continue;
+    for (const block of event.message?.content || []) {
+      if (block?.type === 'text' && typeof block.text === 'string') parts.push(block.text);
+    }
+  }
+  if (result.output) parts.push(result.output);
+  return parts.join('\n');
+}
+
+export function fragmentRestoreVerdict(
+  result: FragmentRestoreRun,
+  markers: { restored: string; others: string[] },
+): { loadedTarget: boolean; didNotLoadOthers: boolean } {
+  const report = agentReport(result);
+  return {
+    loadedTarget: report.includes(markers.restored),
+    didNotLoadOthers: markers.others.every((marker) => !report.includes(marker)),
+  };
+}

test/helpers/free-fixtures-data.ts

@@ -60,6 +60,7 @@ export const FREE_FIXTURES: Readonly<Record<string, readonly string[]>> = {
   'test/fixtures/codex-sandbox/**': ['test/codex-eval-recording.test.ts', 'test/codex-model-probe.test.ts', 'test/outside-review-result.test.ts', 'test/outside-voice-invocation.test.ts'],
   'test/fixtures/context-bill/**': ['test/context-bill.test.ts', 'test/context-budget-ratchet.test.ts'],
   'test/fixtures/context-budget.json': ['test/helpers/capture-context-budget.ts', 'test/test-value-bar.test.ts'],
+  'test/fixtures/context-restore-fragment-match-captures.json': ['test/context-restore-fragment-verdict.test.ts'],
   'test/fixtures/coverage-audit-cd-chain.json': ['test/coverage-audit-evidence.test.ts'],
   'test/fixtures/coverage-audit-parallel-column.json': ['test/coverage-audit-evidence.test.ts'],
   'test/fixtures/coverage-audit-separator-collision.json': ['test/coverage-audit-evidence.test.ts'],

test/helpers/llm-judge.ts

@@ -187,6 +187,13 @@ export async function callJudge<T>(
     .filter(block => block.type === 'text')
     .map(block => block.text)
     .join('\n');
+  const parseJudgeText = (json: string): T => {
+    try {
+      return JSON.parse(json) as T;
+    } catch (error) {
+      throw new SyntaxError(`Judge response is not valid JSON (stop_reason=${response.stop_reason}, ${json.length} chars; full text in llm-judge-response-parse-error): ${error instanceof Error ? error.message : String(error)}`);
+    }
+  };
   try {
     if (response.stop_reason === 'max_tokens') {
       throw new Error(`Judge response truncated at max_tokens=${maxTokens} (model=${resolvedModel})`);
@@ -194,11 +201,11 @@ export async function callJudge<T>(
     if (response.stop_reason === 'refusal') throw new JudgeRefusalError(response);
     if (opts?.jsonSchema !== undefined) {
       if (response.stop_reason !== 'end_turn') throw new Error(`Structured judge did not complete: stop_reason=${response.stop_reason}`);
-      return JSON.parse(text) as T;
+      return parseJudgeText(text);
     }
     const jsonMatch = text.match(/\{[\s\S]*\}/);
     if (!jsonMatch) throw new Error(`Judge returned non-JSON: ${text.slice(0, 200)}`);
-    return JSON.parse(jsonMatch[0]) as T;
+    return parseJudgeText(jsonMatch[0]);
   } catch (error) {
     // The canonical full stderr spool retains this public response even when
     // parsing fails before a caller can record a judgment. Never copy content

test/helpers/touchfiles-data.ts

@@ -682,15 +682,15 @@ export const E2E_TOUCHFILES: Record<string, string[]> = {
   // test/skill-e2e-context-skills.test.ts. These are periodic-tier because
   // each one spawns claude -p and costs ~$0.20-$0.40. Collectively they
   // verify the thing the /checkpoint → /context-save rename was for.
-  'context-save-routing':                  [ 'context-save/**', 'scripts/resolvers/preamble.ts', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts'],
-  'context-save-then-restore-roundtrip':   ['bin/gstack-state-root.sh', 'lib/state-root.ts', 'context-save/**', 'context-restore/**', 'bin/gstack-slug', 'bin/gstack-remote-identity.sh', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts'],
-  'context-restore-fragment-match':        [ 'context-restore/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts'],
-  'context-restore-provenance-order':      ['context-save/**', 'context-restore/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts'],
-  'context-restore-empty-state':           [ 'context-restore/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts'],
-  'context-restore-list-delegates':        [ 'context-restore/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts'],
-  'context-restore-legacy-compat':         [ 'context-restore/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts'],
-  'context-save-list-current-branch':      [ 'context-save/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts'],
-  'context-save-list-all-branches':        [ 'context-save/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts'],
+  'context-save-routing':                  [ 'context-save/**', 'scripts/resolvers/preamble.ts', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts', 'test/helpers/context-restore-fragment.ts'],
+  'context-save-then-restore-roundtrip':   ['bin/gstack-state-root.sh', 'lib/state-root.ts', 'context-save/**', 'context-restore/**', 'bin/gstack-slug', 'bin/gstack-remote-identity.sh', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts', 'test/helpers/context-restore-fragment.ts'],
+  'context-restore-fragment-match':        [ 'context-restore/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts', 'test/helpers/context-restore-fragment.ts'],
+  'context-restore-provenance-order':      ['context-save/**', 'context-restore/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts', 'test/helpers/context-restore-fragment.ts'],
+  'context-restore-empty-state':           [ 'context-restore/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts', 'test/helpers/context-restore-fragment.ts'],
+  'context-restore-list-delegates':        [ 'context-restore/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts', 'test/helpers/context-restore-fragment.ts'],
+  'context-restore-legacy-compat':         [ 'context-restore/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts', 'test/helpers/context-restore-fragment.ts'],
+  'context-save-list-current-branch':      [ 'context-save/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts', 'test/helpers/context-restore-fragment.ts'],
+  'context-save-list-all-branches':        [ 'context-save/**', 'test/skill-e2e-context-skills.test.ts', 'bin/gstack-skill-start', 'test/helpers/shipped-skill-routing.ts', 'test/helpers/context-restore-fragment.ts'],
 
   // Document-release
   'document-release': [ 'test/helpers/outside-voice-evidence.ts', 'bin/gstack-version-bump', 'document-release/**', 'test/skill-e2e-workflow.test.ts', 'test/fixtures/coverage-audit-fixture.ts', 'test/helpers/coverage-audit.ts', 'test/helpers/coverage-audit-evidence.ts', 'test/helpers/office-hours-attempt.ts'],

test/llm-judge-frontier.test.ts

@@ -218,6 +218,22 @@ describe('frontier Claude judge compatibility', () => {
     expect(create).toHaveBeenCalledTimes(1);
   });
 
+  // Train-4 gate 40173759210 slice 4 attempt 1: Sonnet's free-form cookie workflow judgment left the quotes around "one `--domain` per run" unescaped.
+  const COOKIE_GATE_UNESCAPED_REPLY = "{\"clarity\": 4, \"completeness\": 4, \"actionability\": 4, \"reasoning\": \"The workflow is well ordered: Aside check, then consent questions before any `$B` command, then daemon prep, import, and an honest three-line report. It states the boundaries precisely: profile directory vs display name, `--all` exclusions, Chromium-only reset with its non-rollback caveat, and the rule that copied cookies, counts, and HTTP 200 never prove sign-in. The error and failure-reason tables cover recovery and platform limits. Gaps and ambiguities remain. (1) Step 3.1 sets env vars and runs `$B stop` before `$B status` and `goto`, which is workable but slightly awkward. (2) Step 2 says to ask everything before any `$B` command, yet it also needs the Aside probe, which comes earlier in step 1 and so is an implicit exception. (3) It is unclear how the picker or direct import maps the receipt's states to the three report lines, such as `Imported: <count>` for partial or failed results, or the storage-reset line when the picker is used. (4) The `--all` flow is not given a command example or a step. (5) For the multiple-domain case the instruction is \"one `--domain` per run\", but it does not say how to handle the target tab or the receipt across runs. (6) `decryption_failed` says to close the source browser, which sits in tension with the Chrome 136+ note, though it is not contradictory. (7) The `handoff` path requires a display, and the fallback when none is available is not defined beyond 'offer only with a display'. (8) The Windows example's `$B reload` without verification is fine, but the unverified case is not tied to the report. Overall it is strong with a few minor missing decisions.\"}";
+
+  test('the captured cookie judge reply fails closed and names the judge output, stop reason and retained text', async () => {
+    create.mockResolvedValue({ id: 'msg_011CgNbBNCNcZ2C6WJDCmjgU', model: 'claude-sonnet-5-5', stop_reason: 'end_turn',
+      content: [{ type: 'text', text: COOKIE_GATE_UNESCAPED_REPLY }], usage: { input_tokens: 4521, output_tokens: 572 } } as never);
+    let failure: unknown;
+    try { await callJudge('score the workflow', 'claude-sonnet-5-5'); } catch (error) { failure = error; }
+    expect(failure).toBeInstanceOf(SyntaxError);
+    expect((failure as Error).message).toStartWith(`Judge response is not valid JSON (stop_reason=end_turn, ${COOKIE_GATE_UNESCAPED_REPLY.length} chars; full text in llm-judge-response-parse-error): `);
+    expect(create).toHaveBeenCalledTimes(1);
+    expect(diagnostics).toHaveBeenCalledTimes(1);
+    expect(JSON.parse(diagnostics.mock.calls[0][0])).toMatchObject({ type: 'llm-judge-response-parse-error', stopReason: 'end_turn',
+      textBlocks: [COOKIE_GATE_UNESCAPED_REPLY], error: { name: 'SyntaxError', message: (failure as Error).message } });
+  });
+
   test('a weak structured score still fails the substance bar', async () => {
     create.mockResolvedValue({ stop_reason: 'end_turn', content: [{ type: 'text',
       text: JSON.stringify({ reason_substance: 1, reasoning: 'Boilerplate.' }) }] } as never);

test/plan-tune.test.ts

@@ -33,6 +33,7 @@ import {
   validateRegistrySignalKeys,
   newDimensionTotals,
   normalizeToDimensionValue,
+  resolveOptionKey,
   ALL_DIMENSIONS,
 } from '../scripts/psychographic-signals';
 import {
@@ -290,6 +291,125 @@ describe('psychographic signal map', () => {
   });
 });
 
+// -----------------------------------------------------------------------
+// #2623 — chosen UI label → registered option key at capture time
+// -----------------------------------------------------------------------
+
+describe('option label → registered key (#2623)', () => {
+  const words = (key: string) => key.replace(/-/g, ' ');
+  const title = (s: string) => s.charAt(0).toUpperCase() + s.slice(1);
+
+  test('every registered option resolves from its own words, as a bare label and as an AUQ label', () => {
+    const misses: string[] = [];
+    for (const q of Object.values(QUESTIONS as Record<string, QuestionDef>)) {
+      for (const key of (q.options ?? []) as string[]) {
+        for (const label of [key, words(key), `A) ${title(words(key))} (recommended)`, `B. ${title(words(key))}`]) {
+          if (resolveOptionKey(q.id, label) !== key) misses.push(`${q.id}: ${JSON.stringify(label)} → ${resolveOptionKey(q.id, label)}`);
+        }
+      }
+    }
+    expect(misses).toEqual([]);
+  });
+
+  test('common phrasings map to the question\'s own key', () => {
+    const cases: Array<[string, string, string]> = [
+      ['plan-eng-review-arch-finding', 'A) Fix it now (recommended)', 'fix-now'],
+      ['plan-eng-review-arch-finding', 'Fix later', 'defer'],
+      ['plan-eng-review-arch-finding', 'B) Defer to TODOS.md', 'defer'],
+      ['plan-eng-review-arch-finding', 'C) Accept the risk and ship', 'accept-risk'],
+      ['plan-ceo-review-expansion-proposal', 'A) Add to plan (recommended)', 'accept'],
+      ['plan-ceo-review-expansion-proposal', 'B) Add to TODOs', 'defer'],
+      ['plan-ceo-review-mode', 'A) Expand scope (Recommended)', 'expand'],
+      ['plan-ceo-review-mode', 'A) SCOPE EXPANSION', 'expand'],
+      ['plan-ceo-review-mode', 'B) SELECTIVE EXPANSION', 'selective'],
+      ['plan-ceo-review-mode', 'C) HOLD SCOPE (recommended)', 'hold'],
+      ['plan-ceo-review-mode', 'D) SCOPE REDUCTION', 'reduce'],
+      ['plan-ceo-review-mode', 'Cherry-pick expansions', 'selective'],
+      ['plan-ceo-review-mode', 'Cut to the minimum', 'reduce'],
+      ['ship-pre-landing-review-fix', 'Ship as-is', 'skip'],
+      ['ship-test-failure-triage', 'Ship anyway, the failures are known', 'ack-and-ship'],
+      ['review-finding-fix', 'Not a bug, the input is trusted', 'false-positive'],
+      ['land-and-deploy-merge-confirm', 'Yes, merge it', 'accept'],
+      ['land-and-deploy-merge-confirm', 'Don’t merge yet', 'reject'],
+      ['plan-eng-review-test-gap', 'Add the test now', 'add-test'],
+      ['office-hours-mode-goal', 'Hackathon / demo', 'hackathon'],
+    ];
+    for (const [id, label, key] of cases) {
+      expect({ id, label, key: resolveOptionKey(id, label) }).toEqual({ id, label, key });
+    }
+  });
+
+  test('negative controls stay unmapped', () => {
+    // Unregistered (ad-hoc or hook-hash) ids: no registered options to map to.
+    expect(resolveOptionKey('plan-eng-review-ci-typecheck', 'Fix now')).toBeUndefined();
+    expect(resolveOptionKey('hook-0123456789', 'Accept')).toBeUndefined();
+    // Registered question with no options list.
+    expect(resolveOptionKey('plan-devex-review-persona', 'Accept')).toBeUndefined();
+    // A label that names none of the question's options.
+    expect(resolveOptionKey('plan-eng-review-arch-finding', 'Restore the parallel typecheck job')).toBeUndefined();
+    // Cues anchor at the start of the label, never mid-sentence.
+    expect(resolveOptionKey('review-sql-safety', 'Ship it, then fix')).toBeUndefined();
+    expect(resolveOptionKey('review-sql-safety', "Don't fix")).toBeUndefined();
+    // Another question's key is not a candidate.
+    expect(resolveOptionKey('qa-tier', 'Fix now')).toBeUndefined();
+    // A selector-looking number is not stripped into a match.
+    expect(resolveOptionKey('qa-tier', '1.5x deeper')).toBeUndefined();
+    expect(resolveOptionKey('plan-eng-review-arch-finding', '')).toBeUndefined();
+  });
+
+  function logAndRead(entries: Array<Record<string, unknown>>): Array<Record<string, unknown>> {
+    const tmpHome = fs.mkdtempSync(path.join(require('os').tmpdir(), 'gstack-2623-'));
+    try {
+      const { spawnSync } = require('child_process');
+      const env = { ...process.env, GSTACK_HOME: tmpHome, GSTACK_QUESTION_LOG_NO_DERIVE: '1' };
+      const logBin = path.join(ROOT, 'bin', 'gstack-question-log');
+      for (const e of entries) {
+        const r = spawnSync(logBin, [JSON.stringify({ skill: 'plan-eng-review', question_summary: 'q?', ...e })], { env, cwd: ROOT, encoding: 'utf-8', timeout: 30_000 });
+        expect(r.status).toBe(0);
+      }
+      const [slug] = fs.readdirSync(path.join(tmpHome, 'projects'));
+      return fs
+        .readFileSync(path.join(tmpHome, 'projects', slug, 'question-log.jsonl'), 'utf-8')
+        .trim()
+        .split('\n')
+        .map((l) => JSON.parse(l));
+    } finally {
+      fs.rmSync(tmpHome, { recursive: true, force: true });
+    }
+  }
+
+  test('gstack-question-log records the key and keeps the label for hook events on registered ids only', () => {
+    const [hook, agentKey, agentMode, adHoc, unmatched, freeText] = logAndRead([
+      { question_id: 'plan-eng-review-arch-finding', user_choice: 'A) Fix it now (recommended)', recommended: 'Fix it now', source: 'hook' },
+      { question_id: 'plan-eng-review-arch-finding', user_choice: 'defer', recommended: 'fix-now' },
+      { skill: 'plan-ceo-review', question_id: 'plan-ceo-review-mode', user_choice: 'HOLD SCOPE', recommended: 'HOLD SCOPE', auto_decided: true },
+      { question_id: 'plan-eng-review-ci-typecheck', user_choice: 'Fix now', source: 'hook' },
+      { question_id: 'plan-eng-review-arch-finding', user_choice: 'Restore the parallel typecheck job', source: 'hook' },
+      { question_id: 'plan-eng-review-arch-finding', user_choice: 'Fix it but rename the module first', free_text: 'Fix it but rename the module first', source: 'auq-other' },
+    ]);
+    expect(hook).toMatchObject({ user_choice: 'fix-now', choice_label: 'A) Fix it now (recommended)', recommended: 'fix-now', followed_recommendation: true });
+    expect(agentKey).toMatchObject({ user_choice: 'defer', recommended: 'fix-now', followed_recommendation: false });
+    expect(agentKey.choice_label).toBeUndefined();
+    // Agent-written rows keep what the agent logged; the auto-decide harness
+    // matches plan-ceo-review's full mode names.
+    expect(agentMode).toMatchObject({ source: 'agent', user_choice: 'HOLD SCOPE', recommended: 'HOLD SCOPE', followed_recommendation: true });
+    expect(agentMode.choice_label).toBeUndefined();
+    expect(adHoc).toMatchObject({ user_choice: 'Fix now' });
+    expect(adHoc.choice_label).toBeUndefined();
+    expect(unmatched).toMatchObject({ user_choice: 'Restore the parallel typecheck job' });
+    expect(unmatched.choice_label).toBeUndefined();
+    expect(freeText).toMatchObject({ user_choice: 'Fix it but rename the module first', source: 'auq-other' });
+    expect(freeText.choice_label).toBeUndefined();
+  });
+
+  test('a picked label scores as following a key recommendation', () => {
+    const [row] = logAndRead([
+      { question_id: 'plan-ceo-review-mode', user_choice: 'A) Expand scope (Recommended)', recommended: 'expand', source: 'hook' },
+    ]);
+    expect(row).toMatchObject({ user_choice: 'expand', followed_recommendation: true });
+  });
+});
+
 // -----------------------------------------------------------------------
 // Archetypes
 // -----------------------------------------------------------------------

test/qa-only-fixture.test.ts

@@ -19,7 +19,7 @@ test('QA-only fixture helpers and owned sections select the no-fix consumer', ()
 });
 
 test.each(['success', 'max-turns', 'dependencies', 'metadata', 'edit', 'source-write', 'source-delete', 'source-add', 'setup', 'runner', 'recording', 'runner-recording', 'artifact-error',
-  'process-error', 'slow-setup', 'exhausted-setup', 'late-timeout', 'late-retry', 'late-failure-retry', 'unknown-timeout',
+  'process-error', 'slow-setup', 'exhausted-setup', 'late-timeout', 'late-retry', 'late-slow-retry', 'late-failure-retry', 'unknown-timeout',
   'cleanup', 'cleanup-retry', 'runner-cleanup', 'cleanup-recording', 'remove-error', 'server-stop',
   'missing-deadline', 'unguarded-probe', 'forged-receipt', 'deadline-write', 'reset-deadline',
   'preparation-missing', 'preparation-failed', 'preparation-unacknowledged', 'preparation-wrong-path', 'preparation-late',
@@ -47,11 +47,12 @@ if (scenario === 'remove-error') spyOn(fs, 'rmSync').mockImplementation((target,
   return originalRm(target, options);
 });
 mock.module(path.join(root, 'test/helpers/eval-store.ts'), () => ({ getProjectEvalDir: () => ${JSON.stringify(path.join(dir, 'evals'))} }));
-const late = scenario.startsWith('late-'), workMs = 1000, drainMs = 80;
+const late = scenario.startsWith('late-'), lateRetry = ['late-retry', 'late-slow-retry'].includes(scenario), workMs = 1000, drainMs = 80;
 const realNow = Date.now, controlledClock = ['slow-setup', 'exhausted-setup'].includes(scenario);
 const clockStart = realNow();
-let clock = clockStart;
+let clock = clockStart, frozenAt;
 if (controlledClock) Date.now = () => clock;
+else if (late) Date.now = () => frozenAt ?? realNow();
 let finalized = false, outerMs;
 const collector = { addTest: row => {
   const cwd = directories.at(-1);
@@ -77,8 +78,7 @@ mock.module(path.join(root, 'test/helpers/e2e-helpers.ts'), () => ({
   testConcurrentIfSelected: (id, body, budget) => {
     if (id !== 'qa-only-no-fix') return;
     outerMs = budget;
-    const deadline = late ? (budget >= workMs + 5000 ? budget - 5000 + 500 : budget) : 5000;
-    test(id, body, deadline);
+    test(id, late ? () => { frozenAt = realNow(); return body(); } : body, late ? budget : 5000);
   },
   copyDirSync: (source, target) => {
     if (scenario === 'setup') throw new Error('QA-only setup failed');
@@ -87,6 +87,7 @@ mock.module(path.join(root, 'test/helpers/e2e-helpers.ts'), () => ({
   setupBrowseShims: cwd => {
     directories.push(cwd); fs.mkdirSync(path.join(cwd, 'browse/bin'), { recursive: true });
     if (controlledClock) clock += scenario === 'slow-setup' ? 295000 : 300000;
+    if (scenario === 'late-slow-retry') Bun.sleepSync(workMs + 100);
   },
   logCost() {}, createEvalCollector: () => collector, finalizeEvalCollector: async () => { finalized = true; },
   recordE2E: (_collector, name, suite, result, extra) => collector.addTest({ name, suite,
@@ -96,6 +97,7 @@ mock.module(path.join(root, 'test/helpers/e2e-helpers.ts'), () => ({
 mock.module(path.join(root, 'test/helpers/session-runner.ts'), () => ({
   SESSION_DRAIN_GRACE_MS: late ? drainMs : 5000,
   runSkillTest: async options => {
+    frozenAt = undefined;
     const cwd = options.workingDirectory;
     const fact = { cwd, exists: fs.existsSync(cwd), recordedBeforeCapture: records.length,
       reportBefore: fs.existsSync(path.join(cwd, 'qa-reports/qa-only-report.md')), captureMs: options.timeout,
@@ -145,7 +147,7 @@ mock.module(path.join(root, 'test/helpers/session-runner.ts'), () => ({
       fact.existsAfterDrain = fs.existsSync(cwd);
       return { ...result, exitReason: 'timeout' };
     }
-    if (late && (scenario !== 'late-retry' || attempts.length === 1)) {
+    if (late && (!lateRetry || attempts.length === 1)) {
       await new Promise(resolve => setTimeout(resolve, options.timeout + drainMs));
       fact.existsAfterDrain = fs.existsSync(cwd);
       return { ...result, exitReason: 'timeout' };
@@ -237,7 +239,7 @@ await import(path.join(root, 'test/skill-e2e-qa-workflow.test.ts'));
       expect(records[index].cleanupDone).toBe(true);
       expect(cleanups[index].budget).toBeGreaterThan(0);
       expect(cleanups[index].budget).toBeLessThanOrEqual(4000);
-      expect(records[index].passed).toBe(scenario === 'recording' || !shouldFail && !(['late-retry', 'cleanup-retry'].includes(scenario) && index === 0));
+      expect(records[index].passed).toBe(scenario === 'recording' || !shouldFail && !(['late-retry', 'late-slow-retry', 'cleanup-retry'].includes(scenario) && index === 0));
       if (scenario.includes('cleanup') && !(scenario === 'cleanup-retry' && index === 1)) expect(records[index].error).toContain('QA-only browser cleanup: invalid daemon identity');
     }
     for (const [index, attempt] of attempts.entries()) {
@@ -257,6 +259,7 @@ await import(path.join(root, 'test/skill-e2e-qa-workflow.test.ts'));
     }
     if (scenario.startsWith('late-')) {
       expect(outerMs).toBe(1000 + 80 + 5000);
+      for (const attempt of attempts) expect(attempt.captureMs).toBe(1000);
       expect(records[0]).toMatchObject({ passed: false, exit_reason: 'timeout', cost_usd: 0.25 });
       expect(records[0].error).toContain('timeout');
       if (count === 2) expect(directories[0]).not.toBe(directories[1]);

test/session-update-autostash.test.ts

@@ -1,5 +1,5 @@
 import { afterEach, describe, test, expect } from 'bun:test';
-import { execFileSync, spawnSync } from 'child_process';
+import { execFileSync, spawn, spawnSync } from 'child_process';
 import * as fs from 'fs';
 import * as os from 'os';
 import * as path from 'path';
@@ -25,7 +25,7 @@ function makeFixture() {
   const install = path.join(base, 'install');
   const state = path.join(base, 'state');
   fs.mkdirSync(state, { recursive: true });
-  execFileSync('git', ['init', '-q', '--bare', '-b', 'main', origin]);
+  execFileSync('git', ['init', '-q', '--bare', '-b', 'main', origin], { timeout: 30_000 });
 
   fs.mkdirSync(path.join(seed, 'bin'), { recursive: true });
   fs.writeFileSync(path.join(seed, 'VERSION'), '1.0.0\n');
@@ -45,14 +45,22 @@ function makeFixture() {
   git(seed, 'branch', '-M', 'main');
   git(seed, 'remote', 'add', 'origin', origin);
   git(seed, 'push', '-q', 'origin', 'main');
-  execFileSync('git', ['clone', '-q', origin, install]);
+  execFileSync('git', ['clone', '-q', origin, install], { timeout: 30_000 });
   return { base, origin, seed, install, state };
 }
 
-function runScript(install: string, state: string) {
+// Each fixture owns its state root (GSTACK_STATE_ROOT tops the chain in
+// bin/gstack-state-root.sh) and its workflow-lease directory. The preload's
+// GSTACK_LEASE_DIR is shared by every test file in this Bun process, so a lease
+// another file leaves there made the updater log SKIP leased (#3385).
+function fixtureEnv(base: string, install: string, state: string) {
+  return { ...process.env, GSTACK_DIR: install, GSTACK_STATE_ROOT: state, GSTACK_LEASE_DIR: path.join(base, 'leases') };
+}
+
+function runScript(base: string, install: string, state: string) {
   return spawnSync('bash', [SCRIPT], {
     encoding: 'utf8',
-    env: { ...process.env, GSTACK_DIR: install, GSTACK_STATE_DIR: state },
+    env: fixtureEnv(base, install, state),
     timeout: 20000,
   });
 }
@@ -85,7 +93,7 @@ describe('gstack-session-update pull wedge (#2566)', () => {
       // prefix-rename / gbrain-block shape: tracked file, modified).
       fs.appendFileSync(path.join(install, 'SKILL.md'), 'locally patched line\n');
 
-      const r = runScript(install, state);
+      const r = runScript(base, install, state);
       expect(r.status).toBe(0);
       const log = await waitForLog(state, /UPDATING|UP_TO_DATE|PULL_FAILED/);
       expect(log).not.toContain('PULL_FAILED');
@@ -110,7 +118,7 @@ describe('gstack-session-update pull wedge (#2566)', () => {
       git(seed, 'commit', '-aqm', 'remote divergence');
       git(seed, 'push', '-q', 'origin', 'main');
 
-      const r = runScript(install, state);
+      const r = runScript(base, install, state);
       expect(r.status).toBe(0);
       const log = await waitForLog(state, /PULL_FAILED/);
       expect(log).toContain('PULL_FAILED');
@@ -123,6 +131,70 @@ describe('gstack-session-update pull wedge (#2566)', () => {
   }, 30000);
 });
 
+describe.skipIf(process.platform === 'win32')('gstack-session-update fixtures ignore leases outside their own (#3385)', () => {
+  const LEASE = path.join(ROOT, 'bin', 'gstack-lease');
+  const sleepers: ReturnType<typeof spawn>[] = [];
+  afterEach(() => { for (const s of sleepers.splice(0)) s.kill('SIGKILL'); });
+
+  function holdLease(install: string, leaseDir: string): number {
+    const sleeper = spawn('sleep', ['120'], { stdio: 'ignore' });
+    sleepers.push(sleeper);
+    const touch = spawnSync('bash', [LEASE, 'touch', '--root', install, '--pid', String(sleeper.pid)], {
+      encoding: 'utf8', env: { ...process.env, GSTACK_LEASE_DIR: leaseDir }, timeout: 10_000,
+    });
+    expect(touch.status, touch.stderr).toBe(0);
+    return sleeper.pid!;
+  }
+
+  function advanceUpstream(seed: string): string {
+    fs.writeFileSync(path.join(seed, 'VERSION'), '1.1.0\n');
+    git(seed, 'commit', '-aqm', 'upstream release');
+    git(seed, 'push', '-q', 'origin', 'main');
+    return git(seed, 'rev-parse', 'HEAD');
+  }
+
+  test('a live lease in the shared preload lease directory does not stop the fixture update (negative control)', async () => {
+    const shared = process.env.GSTACK_LEASE_DIR!;
+    expect(shared).toBeTruthy();
+    const { base, seed, install, state } = makeFixture();
+    let pid = 0;
+    try {
+      const upstreamHead = advanceUpstream(seed);
+      pid = holdLease(install, shared);
+      const ambient = spawnSync('bash', [LEASE, 'live', '--root', install], { encoding: 'utf8', env: process.env, timeout: 10_000 });
+      expect(ambient.status).toBe(0);
+      expect(ambient.stdout).toContain(`${pid} `);
+
+      expect(runScript(base, install, state).status).toBe(0);
+      const log = await waitForLog(state, /UPDATING|PULL_FAILED|SKIP /);
+      expect(log).not.toContain('SKIP leased');
+      expect(log).toContain('UPDATING');
+      expect(git(install, 'rev-parse', 'HEAD')).toBe(upstreamHead);
+    } finally {
+      if (pid) fs.rmSync(path.join(shared, String(pid)), { force: true });
+      fs.rmSync(base, { recursive: true, force: true });
+    }
+  }, 30000);
+
+  test("the fixture's own live lease still skips the update", async () => {
+    const { base, seed, install, state } = makeFixture();
+    try {
+      const before = git(install, 'rev-parse', 'HEAD');
+      advanceUpstream(seed);
+      const pid = holdLease(install, path.join(base, 'leases'));
+
+      expect(runScript(base, install, state).status).toBe(0);
+      const log = await waitForLog(state, /SKIP leased /);
+      expect(log).toMatch(new RegExp(`SKIP leased pid=${pid}/\\d+m`));
+      expect(log).not.toContain('UPDATING');
+      expect(git(install, 'rev-parse', 'HEAD')).toBe(before);
+    } finally {
+      fs.rmSync(base, { recursive: true, force: true });
+    }
+  }, 30000);
+});
+
+
 // ── #2613: the lock pidfile must record the LIVE holder, not the exited parent ──
 //
 // `echo $$` inside the backgrounded subshell recorded the parent hook's PID.
@@ -144,10 +216,10 @@ describe('gstack-session-update lock identity + TTL (#2613)', () => {
     return shimDir;
   }
 
-  function runScriptWithPath(install: string, state: string, shimDir: string) {
+  function runScriptWithPath(base: string, install: string, state: string, shimDir: string) {
     return spawnSync('bash', [SCRIPT], {
       encoding: 'utf8',
-      env: { ...process.env, GSTACK_DIR: install, GSTACK_STATE_DIR: state, PATH: `${shimDir}:${process.env.PATH}` },
+      env: { ...fixtureEnv(base, install, state), PATH: `${shimDir}:${process.env.PATH}` },
       timeout: 20000,
     });
   }
@@ -160,7 +232,7 @@ describe('gstack-session-update lock identity + TTL (#2613)', () => {
     const { base, install, state } = makeFixture();
     const shimDir = makeSlowGitShim(base, 3);
     try {
-      const r = runScriptWithPath(install, state, shimDir);
+      const r = runScriptWithPath(base, install, state, shimDir);
       expect(r.status).toBe(0); // parent hook has EXITED by now (spawnSync waited)
       // Poll for the pidfile the detached subshell writes.
       const pidPath = path.join(state, '.setup-lock', 'pid');
@@ -191,7 +263,7 @@ describe('gstack-session-update lock identity + TTL (#2613)', () => {
       const lockDir = path.join(state, '.setup-lock');
       fs.mkdirSync(lockDir, { recursive: true });
       fs.writeFileSync(path.join(lockDir, 'pid'), String(holder.pid));
-      const r = runScript(install, state);
+      const r = runScript(base, install, state);
       expect(r.status).toBe(0);
       const log = await waitForLog(state, /SKIP locked_by=/);
       expect(log).toContain(`SKIP locked_by=${holder.pid}`);
@@ -209,7 +281,7 @@ describe('gstack-session-update lock identity + TTL (#2613)', () => {
       const lockDir = path.join(state, '.setup-lock');
       fs.mkdirSync(lockDir, { recursive: true });
       fs.writeFileSync(path.join(lockDir, 'pid'), String(dead.pid));
-      const r = runScript(install, state);
+      const r = runScript(base, install, state);
       expect(r.status).toBe(0);
       const log = await waitForLog(state, /UP_TO_DATE|UPDATING/);
       expect(log).toMatch(/UP_TO_DATE|UPDATING/);
@@ -224,7 +296,7 @@ describe('gstack-session-update lock identity + TTL (#2613)', () => {
       const lockDir = path.join(state, '.setup-lock');
       fs.mkdirSync(lockDir, { recursive: true });
       fs.writeFileSync(path.join(lockDir, 'pid'), ''); // mkdir→echo race window
-      const r = runScript(install, state);
+      const r = runScript(base, install, state);
       expect(r.status).toBe(0);
       const log = await waitForLog(state, /SKIP locked_by=/);
       expect(log).toContain('SKIP locked_by=');
@@ -290,7 +362,7 @@ describe('gstack-session-update lock identity + TTL (#2613)', () => {
       // forever, so liveness alone can never clear this wedge.
       const past = new Date(Date.now() - 40 * 60 * 1000);
       fs.utimesSync(pidPath, past, past);
-      const r = runScript(install, state);
+      const r = runScript(base, install, state);
       expect(r.status).toBe(0);
       const log = await waitForLog(state, /RECLAIMED lock_ttl_expired/);
       expect(log).toContain('RECLAIMED lock_ttl_expired');

test/session-update-stages.test.ts

@@ -364,5 +364,20 @@ describe.skipIf(process.platform === 'win32')('gstack-session-update honors work
     const log = await waitFor(dead, /UPDATED from=1\.0\.0 to=1\.1\.0/);
     expect(log).not.toContain('SKIP leased');
   }, 30_000);
+
+  test("the starting session's own lease never blocks its update, behind a shell wrapper too", async () => {
+    const fx = makeFixture();
+    upstreamRelease(fx, '1.1.0');
+    touchLease(fx, process.pid);
+    const r = spawnSync('/bin/sh', ['-c', 'bash "$0"; exit $?', SCRIPT], {
+      encoding: 'utf8',
+      env: { PATH: `${path.dirname(process.execPath)}:${process.env.PATH}`, HOME: fx.home, GSTACK_DIR: fx.install, GSTACK_STATE_ROOT: fx.state, SETUP_CONTROL: fx.control, SETUP_CALLS: fx.calls, TMPDIR: fx.base, ...leaseEnv(fx) },
+      timeout: 20_000,
+    });
+    expect(r.status, r.stderr).toBe(0);
+    const log = await waitFor(fx, /UPDATED from=1\.0\.0 to=1\.1\.0/);
+    expect(log).not.toContain('SKIP leased');
+    expect(setupRuns(fx)).toBe(1);
+  }, 30_000);
 });
 

test/skill-e2e-context-skills.test.ts

@@ -23,6 +23,7 @@ import {
 } from './helpers/e2e-helpers';
 import { extractSkillBody } from './helpers/skill-fixture';
 import { readShippedSkillRouting } from './helpers/shipped-skill-routing';
+import { fragmentRestoreVerdict } from './helpers/context-restore-fragment';
 import { spawnSync } from 'child_process';
 import * as fs from 'fs';
 import * as path from 'path';
@@ -260,11 +261,14 @@ Do NOT use AskUserQuestion.`,
 
     logCost('context-restore-fragment-match', result);
 
-    // Broader surface — agent may stop at Skill call without echoing the
-    // body marker. The payments file's body is in tool outputs (Read/Bash).
-    const out = fullOutputSurface(result);
-    const loadedPayments = out.includes('FRAGMATCH_PAYMENTS_BUILD');
-    const didNotLoadOthers = !out.includes('FRAGMATCH_ALPHA_BUILD') && !out.includes('FRAGMATCH_OMEGA_BUILD');
+    // What was restored is what the agent reports, not what its tools printed:
+    // finding the matching title can cat every checkpoint, so all three bodies
+    // may appear in Bash/Read output. The report must carry the payments body
+    // and no other checkpoint's body.
+    const { loadedTarget: loadedPayments, didNotLoadOthers } = fragmentRestoreVerdict(result, {
+      restored: 'FRAGMATCH_PAYMENTS_BUILD',
+      others: ['FRAGMATCH_ALPHA_BUILD', 'FRAGMATCH_OMEGA_BUILD'],
+    });
     const routedToRestore = skillCalls(result).includes('context-restore');
     const exitOk = ['success', 'error_max_turns'].includes(result.exitReason);
 

test/skill-e2e-design.test.ts

@@ -1316,8 +1316,8 @@ describeIfSelected('Design HTML slop gate E2E', ['design-html-slop-gate'], () =>
   testConcurrentIfSelected('design-html-slop-gate', async () => {
     const result = await runSkillTest({
       prompt: `Read design-html-gate.md: the /design-html detector probe block and its "Slop Gate (bounded, never a loop)" step.
-finalized.html in this directory is the finished page. Run the probe (--host claude), then the slop gate on finalized.html exactly as written: one surgical fix pass over the non-advisory findings, one rescan, then stop.
-Write ${workDir}/gate-output.md listing what you fixed and every remaining finding as accepted-with-reason, each tagged with its [rule-id]. Do not take screenshots, do not run npx, do not scan more than twice.`,
+finalized.html in this directory is the finished page. Run the probe (--host claude), then the slop gate on finalized.html exactly as written: one surgical fix pass over the non-advisory findings, one rescan. Then the craft check line that follows it, the same way, then stop.
+Write ${workDir}/gate-output.md listing what you fixed and every remaining finding as accepted-with-reason, each tagged with its [rule-id] or craft id. Do not take screenshots, do not run npx, do not run either check more than twice.`,
       workingDirectory: workDir,
       maxTurns: 20,
       timeout: CAPTURE_MS,
@@ -1332,6 +1332,9 @@ Write ${workDir}/gate-output.md listing what you fixed and every remaining findi
     const scans = result.toolCalls.filter(c => c.tool === 'Bash' && /gstack-design-detect\.ts scan /.test(String(c.input?.command ?? '')));
     expect(scans.length).toBeGreaterThanOrEqual(1);
     expect(scans.length).toBeLessThanOrEqual(2);
+    const craftRuns = result.toolCalls.filter(c => c.tool === 'Bash' && /gstack-design-craft['"]?\s+[^\s;|&]*\.html\b/.test(String(c.input?.command ?? '')));
+    expect(craftRuns.length).toBeGreaterThanOrEqual(1);
+    expect(craftRuns.length).toBeLessThanOrEqual(2);
     const outPath = path.join(workDir, 'gate-output.md');
     expect(fs.existsSync(outPath)).toBe(true);
     const out = fs.readFileSync(outPath, 'utf-8').toLowerCase();

test/test-of-test-ratchet.test.ts

@@ -60,6 +60,7 @@ const BASELINE = [
   'test/codex-carve-fixture.test.ts',
   'test/codex-eval-recording.test.ts',
   'test/codex-resume-flag-semantics.test.ts',
+  'test/context-restore-fragment-verdict.test.ts', // GSTA-91: replays three captured false reds (0/3 on main cb99df4) through the paid context-restore-fragment-match case's fragmentRestoreVerdict
   'test/cookie-workflow-judge-input.test.ts', // slices test/skill-llm-eval.test.ts source (recorded 2026-10 when the ratchet learned paid-source slicing)
   'test/cso-ntfs-fixture.test.ts',
   'test/cso-windows-build-contract.test.ts',