Stronghold successor review native replay
The repaired same-command publication barrier passed in the sole native attempt. The original physical route reached ChooseType, completed fresh Custom Game hover review and its balanced press, then the original browser deadline closed Chrome before setup arrival. Preset, native Start, world and visible gameplay remain unverified.
This assignment uses the tested consumed-request publication repair from
the reviewed native predecessor,
commit 943dce309, with the exact hover route from
preset hover recognition,
commit 2b5b3b07e. Fresh evidence is confined to
scratch/runs/20261104-stronghold-hover-reviewed-native-world/successor-native/;
the predecessor's sealed files remain intact.
The allocation is bx_mxpzefha, expiry 2026-11-04T14:00:53.847Z,
guest cutoff 13:48Z, recovery by 13:53Z and coordinator shutdown
checkpoint 13:57Z. The worker leaves the box running and performs no
lifecycle operation. Original cumulative navigation16384, route4096,
arrival3072, stability768, ACK64, absolute command40 seconds and browser360
seconds remain unchanged. The predecessor consumed14336 startup steps;
that observation grants no larger allowance or deadline reset.
Preflight explicitly resolves all36 predecessor dependencies, replacing the
nine runtime files, operator and state helper, and publication helpers with
the assignment's recorded identities. Authorized allocation-bound transport,
release and cache checker make39 staged files. The repaired publication helper
is intentionally different from the old closure identity. hover-layout.json
remains 0632e31b8c0a62be221de882e019aa1dedcca512a983436db548e719f60ab318.
An initial staging-path mapping error was caught before Chrome and corrected;
the runtime helpers were unchanged.
All2415 original pins passed: runtime source
1188e9bf57ef35bd3363fb88b7ca0da442ce696d, canonical WASM
cffca869b541e6eff12ae3a6b74c4476977e0a7e23cd43d8ec5ba8faecd4bcee, companion
580f49f3c78d26e6855efa0beb37046d03ea24690ff98cfcb3e94ce5e6cb1ee3.
Actual wrapper setup and lazy imports, Puppeteer CJS/ESM, executable access,
98 page/12 Worker cache entries and five pinned HTTP isolation responses
passed without creating a browser. Module and media paths are referenced;
no module/media copies or source archives are made.
Historical RGBA was restored inside this assignment's evidence directory from retained adjacent gzip originals. Both original SHA identities and the coordinator returned-evidence archive roundtrip manifests were verified. Compressed originals remain intact. Historical captures and offline fixtures grant no live review or input permission.
Before Chrome, fresh tests exercised the staged identities:25 actual-driver cases,25 focused layout negatives, actual driver/wrapper/separate fetch/ACK transport integration,18 operator controls, six personal-review controls, four transport controls, six split controls, nine bounded-route controls, three positive/48 negative retained layout controls, actual operator entry, exclusive lock, timely reply inspection and actual wrapper top-level setup. The actual two-review wrapper regression reproduced the original consumed request return and passed with the repaired helper. Four additional cases covered never-republished consumed request, future binding with consumed ID, malformed consumed ID and another command's history. All ten retained publication controls passed, including an actual40.278-second timeout. These offline fixtures granted no live permission. Full RGBA and independent column recognizers remain exact, with no mask, tolerance or foreign layout.
Chrome15183 recorded identity at 13:28:24.693Z, with the original absolute deadline 13:34:18.464Z. Input-free commands0–6 completed and their successful replies were inspected in time. Captures2048/4096 show the loading artwork, 6144 shows additional media loading,8192/10240 show the loading artwork, and 12288 already shows Enter Your Title. The worker submitted the next2048-step observation in the same tool call that retrieved the12288 image, before personally interpreting that image. This avoidable sequencing error consumed 2048 additional navigation steps and wall time;14336 was used before title input. No counter was reset or allowance extended to compensate.
Command7 used current fetch, personal VIEW and a separate ACK for the title, then a distinct current fetch, personal VIEW and separate ACK for its main menu. Balanced Enter consumed128 steps. The repaired helper did not return the acknowledged title request; command7's final success was inspected before its original deadline. Command8 independently reviewed the main menu and Crusader hover under the same command index, then completed the balanced Crusader press, using192 steps. Its final success was separately inspected. These two actual multi-review commands establish the live publication repair.
Command9 personally reviewed original ChooseType and installed the single
route. Command10 separately reviewed ChooseType. Command11 emitted only the
Custom Game MOVE, fetched the fresh hover request, received personal VIEW
and separate ACK, and completed its timely inspected success. Command12
executed the balanced Custom Game press and returned continuation: arrival
at 13:34:12.540Z, with320 cumulative route steps and14976 navigation
steps. There was no arrival or stability command, setup review, preset MOVE,
preset hover review/press, handoff or native Start. The final recovered route
capture was personally viewed: it still shows ChooseType with Custom Game
hover artwork, rather than setup or gameplay. That recovery inspection grants
no subsequent input permission.
Only seconds remained for the required dependent setup/stability/review
sequence. Quit13 was submitted NEXT, but Chrome had already closed normally
at 13:34:19.134Z, exit0, for bounded browser deadline. Closure was354.441
seconds after browser identity and670ms after the timer deadline; the original
360-second interval was not renewed. Quit13 consequently had no reply and the
operator retained its original timeout latch. Commands0–12 all had timely
inspected successful replies. There was no earlier driver failure or second
Chrome, retry, reload, forced callback, guest write, repaint or fault clearing.
actual-result.json verifies234 complete64-step ACKs,14976 steps,1032 matching
served pins, seven personal reviews and drained final input. Captures are
original host PNGs; no live guest raw-pixel dump or derived PNG is claimed.
The restored historical RGBA is explicitly historical. Neither those pixels
nor the new screenshots establish native world state or visible gameplay.
All315 immutable closed artifacts/19457914 bytes and23 unique control files
were recovered with full SHA verification. A second control inventory was
identical. Final2415 pins passed, Chrome and the replay process were absent,
and HTTP46467 was closed. Custody returned at 13:35Z with bx_mxpzefha
RUNNING (API idle, healthy and fully hydrated), expiry
2026-11-04T14:00:53.847Z. The coordinator retains normal shutdown at13:57Z;
this worker performs no lifecycle action or further runtime work after return.
The coordinator subsequently reported normal stop completion at 13:37:36Z,
forced: false, error: null, in its13:38 messageboard entry, citing
coordinator-returned-boxes-1338.json. That later lifecycle result is the
coordinator's report; this worker did not stop or requery the returned box.
The tested next recipe retains the repaired publication helper and exact hover closure. For any separately authorized allocation, regenerate transport, release and all staged identities, then rerun the actual controls before its sole browser. Inspect each observation's reply and personally VIEW its image in a completed tool call before deciding whether another observation is needed. At the first actual title capture, use the unchanged title/menu route without an extra observation. Each multi-review command still needs fresh current fetch → personal VIEW → separate ACK for every distinct request, then timely final reply inspection. No helper repair or larger bound is justified by this worker's observation sequencing error.
After ChooseType and Custom Game, the unchanged bounded arrival and three
stability captures must precede fresh setup review. Only then use MOVE-only
route-next, separate route-review-preset-hover with its own current
fetch/VIEW/ACK and timely inspected success, then separate balanced
route-press-preset. Fresh Zephyr review/handoff and unchanged native-Start
guards remain mandatory. Stop at the first failed prerequisite, allowing only
quit NEXT; do not reset counters or extend any original deadline. A future
world claim still requires actual native state and a personally viewed fresh
screenshot, with raw pixels retained before any derived PNG.
Git identity, relevant binary diff, small run-helper/note patches and an artifact seal preserve provenance. Only this note is committed. No production code changed and no full-build result is claimed.