Scoped negative result: exact 73-primitive factory scale optimum (no new kappa)

research/complex-factory-scale-frontier27/CLAIM.json

@@ -0,0 +1,51 @@
+{
+  "track": "Complex supplier",
+  "assignment_seed": 27,
+  "randomness": "Python secrets.randbelow(100), operating-system randomness",
+  "base_commit": "47481c89fd30a43f8f882eb4b391e89c29273fd0",
+  "started_utc": "2026-10-13T13:51:45Z",
+  "claim_utc": "2026-10-13T13:58:46.626396+00:00",
+  "expected_finish_utc": "2026-10-13T14:31:45Z",
+  "verification_handoff_reserved_from_utc": "2026-10-13T14:21:45Z",
+  "coordination": "GitHub pinnedIssues empty; only open issues 913 and 650, neither a sprint coordination issue. Local claim only.",
+  "object": "PR922/PR924 combined 26-addition raw-factory topology, all signed dyadic scale placements under exactly 32 scalar macros",
+  "method": "Reduce minimum-macro gauge class to one fixed net scale on each of six wires. Enumerate all placements modulo disjoint-wire commutation, using seed 27 to order wires/ties. Verify exact matrices, all orientations, complete primitive costs and prefix consequences.",
+  "hypothesis": "A better placement may lower the 73-primitive witness without adding macros; otherwise certify the exact restricted optimum.",
+  "deliverable": "Source-pinned exact census and witness, proof of scope reduction, executable verifier, negative controls, one research PR; no new kappa.",
+  "candidates": [
+    {
+      "task": "Four-to-three factory contractions",
+      "decision": "Not chosen: recent PR922 contraction work; harder to separate scope within bounded session."
+    },
+    {
+      "task": "CK helper-run coefficient reuse",
+      "decision": "Not chosen: PR769/860/919 overlap heavily with this mechanism."
+    },
+    {
+      "task": "Minimum-macro signed-scale placement frontier for contracted factory",
+      "decision": "Chosen: PR924 explicitly leaves macro-budget optimality open. Uses new fixed 26-addition topology and complete finite classification, not a new random seed on an old search."
+    }
+  ],
+  "reviewed_overlap": {
+    "922": "efd5643c0016b6c6dac26e9ed7dee8a75295ff2a",
+    "924": "d02d83332581f6a23616c1366c6b3901d42f4b06",
+    "931": "bounded alternate butterfly topology, distinct class",
+    "919": "CK helper sign normalization, distinct raw word"
+  },
+  "recent_pr_scan_through": 945,
+  "attribution": "Ross Richardson with substantial OpenAI Codex assistance; no delegated agents.",
+  "result_update": {
+    "updated_utc": "2026-10-13T14:19:06.349646+00:00",
+    "status": "Completed scoped negative result; ready for one focused PR",
+    "finding": "Exactly 73 primitives is the minimum at 32 macros for the fixed 26-ADD signed-dyadic gauge class; 39,424 magnitude and 315,392 complete placements, one magnitude optimum/eight sign ties. Extra macros cannot lower the retained complete local invoice.",
+    "new_kappa": false,
+    "reviewable_branch_link": "https://github.com/CrocSwap/integer-mult-bounds/compare/main...RossRichardson:integer-mult-bounds:research/complex-supplier-seed27",
+    "reproduction": "python3 -I -O -B research/complex-factory-scale-frontier27/check.py",
+    "verification": "CPython 3.14.6 and 3.9.6 full package pass; exact independent nine-column replay of all 39,424 magnitude representatives; 12 test groups; selected-record check passes.",
+    "refreshed_main": "007664306fff84fbb1c9a1b4053b69f439764034",
+    "final_overlap_inventory_count": 700,
+    "final_overlap_scan_through": 947,
+    "dependency_heads_unchanged": true,
+    "no_pinned_coordination_issue": true
+  }
+}