[Jammy2211]

Part of #499 (PR 1 of 2; the PyAutoPulse day_zero snapshot follows once this merges). Likelihood lifecycle phase 3, day-zero write 2/4.

Heart RED override: invoked by the human on 2026-11-01 (development only). Tier: judge, so this merges only by the human's /prm.

What changes

• catalogue/day_zero.json (schema day-zero v1) declares the frozen baseline:

  • tag 2026.10.25.1;
  • the §3.6 rule;
  • the commits it was frozen against: autolens_profiling 8c40e7d (after feat(catalogue): class records from the Pulse designations; retire provisional_designations.json #498) and the PyAutoPulse designations bcaf6b86 (designations: record the 39 phase-3 day-zero designations (#83) PyAutoPulse#84);
  • the human's rulings.

• build_catalogue stamps day_zero: "2026.10.25.1" on every current record of a cell that meets the rule:

  • class optimal or secondary;
  • regime standard;
  • reference host (hpc_ral_cpu / hpc_a100);
  • release_pin on the tag;
  • a current parity match and a fit figure.

  A passing cell is stamped whole. Stamped records' validation reason names the baseline; validation.status stays unreviewed.

• validate_lifecycle refuses any stamp that breaks the rule.

Result

• 480 records over 116 cells are stamped. Nothing was re-measured.
• All 39 designations have a stamped current optimal.
• The held datacube/delaunay A100 group (20 records) is included, per the human's ruling.
• Experimental rows, the 784-pixel size-variant rows and rows still on 2026.8.17.1 stay unstamped.

Checks run locally

• build_catalogue.py --check --validate-with ../PyAutoPulse, coverage_matrix.py --check, build_dashboard.py --check, build_readme.py --check, check_wiki.py --check: OK.
• ruff check and ruff format --check: clean.
• pytest scripts/misc/*/test: no new failures against main. New tests cover a stamped cell, three unstamped cases (off the tag, parity diverged, no figure), a refused stamp, and the real-catalogue witness (480 records, 116 cells, 39 designations, 20 held).

---

Generated by Claude Code