The first full local ./buddy test run while qualifying the host-slot coordinator failed the existing test camera uncertainty affects comparability and missing stations cannot masquerade as completed collection in tests/unit/inspection-evidence.test.ts:122.
The runner reports a 30,000 ms timeout and an observed duration of 134,383.85 ms. No failed assertion, intra-test timing or stack identifying the stalled operation appears in that failure. The original suite is still running, so this is an initial observation, not a terminal suite result. Runtime is PATH Pantry Bun 1.4.3 on macOS. Source HEAD is 222ffb83c8a3ae934eea339d7451b37b69bbc1b7 with the separate uncommitted host-slot prototype.
All eleven inspection source/test files in the retained snapshot byte-match that published HEAD. This establishes source identity, not the cause or proof that other changes cannot affect the run. The case exercises real owned simulation, uncertainty comparison, missing-station rejection and incomplete-record validation. No acceptance assertion or deadline has been changed, and no retry has occurred.
Retained initial log prefix: 27431 bytes, SHA-256 eff851537599640faea11a7a392ed0da5eee36441754e4498839a1099b0a3b1f. The original full log continues separately. Do not infer host sleep, CPU load, memory pressure, a Bun issue or algorithm failure from the duration alone.
Acceptance:
- Retain the original terminal suite result and all other failures separately.
- Run a bounded focused diagnostic against the actual original case, retaining every verdict and bounded operation timing. A passing repeat alone does not explain the original failure.
- Establish a cause before applying a production or fixture correction. Preserve uncertainty bounds, missing-station rejection and the existing deadline unless independent requirements justify a change.
- Qualify any correction with the relevant inspection/full suite and exact-head CI. Keep physical inspection and hardware evidence distinct.
Refs #20. Host-slot publication remains unqualified pending its full-suite assessment; this ticket does not close the new coordinator or inspection acceptance.
Original terminal result
The original full local Buddy run completed with exit 1: 1,502 passed, 25 skipped and three failed, 489,058 assertions, 1,530 tests across 187 files, 3,683.77 seconds. The failures are the inspection timeout (#52), 137-review 404 (#36), and purge-seal child exit 137 (#53). The host-slot/runtime cases passed. These failures remain separate observations with unproven causes.
The retained full terminal log has 427362 bytes and SHA-256 6d884c7966f217c823a20cc439b8636035c99b160b648ea6e92ced5e5e29ab7c. No original deadline, assertion or production policy was weakened, and this result is not a passing full suite.
Bounded follow-up
After the original suite terminated, the unchanged original case passed, and the entire unchanged inspection-evidence file passed separately, both retaining the 30,000 ms per-test limit under the same PATH Bun 1.4.3. Four supplemental fresh-process repetitions retained the original assertions and actual simulation/comparison/validation calls, adding only private phase timing and CPU/RSS records. All four passed; their summed measured operation times were 3660.5 ms, 4468.3 ms, 4620.09 ms, 4346.19 ms. Every verdict and raw phase record is retained. No product or test source was changed for these operations.
This is a non-reproduction, not a root-cause fix or proof of arbitrary-host timing. The original full-suite timeout remains unexplained; this issue stays open.
Tracker correction
A negated closing phrase in an earlier PR description created an unintended automatic closing reference. The phrase has been removed. This issue is reopened because the original cause remains unresolved; no causal correction has been established. The passing follow-up and improved observations retain their documented limits.