Pilot preflight, 2026-10-14 12:37 Halifax. Read-only checks made before the pilot's first piece, as Colson asked with his go (Oct 14). These checks change nothing.
Freeze record: medium_freeze.json, made 2026-10-14 12:29 Halifax.

1. Fingerprints (SHA-256) of the plan, the code and the calibration's results, frozen against now:
   plan: b2592cfea01be207022145da51d78b48c9f30dc1ff5dc53c0ae86026abe7f427: matches
   code: 6b75b7801692298dff0000f1e5fbebdae18285a8b1fedce8805f616e286b793e: matches
   calibration_json: 7a1c24f636a159d450c45b04d5285fe0ba129e51c123fc6ecc695bbb893ed7b9: matches
   calibration_txt: f0f9d1987ac35f1615e372dbfc65c1eafc6589157313954782429a0cc58f140f: matches
2. The code's settings as frozen: yes
3. Protected folders and files (the cause test with its FAIL record, its exploration and prelock work including the paused hold-versus-pulse design, the nudge test, the coherence-growth folder, the cancer-research files outside this test, and plan-draft-1.md): 103 files, all matching the freeze: yes
4. The calibration's saved parts: 42 files, all matching the freeze: yes
5. The cause test's locked files (the plan, code and inputs its result depends on) unchanged: yes
6. No pilot output exists yet (no pilot folder, run log or results): yes
7. The pilot's starts (103 to 114) and their moment times as frozen: yes
8. The size the pilot will read from the freeze: 0.5% of K (K = 1.9095 raised, 1.8905 lowered, then exactly 1.9): yes
9. The same numerical environment as the calibration: yes
10. The pilot's own freeze check (run again before every piece and before the report): passes

Everything as frozen and no pilot output yet: yes