fix(benchmarks): address hadamard-order12 verifier review comments

- Change evidence media type from text/plain to application/json with
  payload_shape in instruction.md and public_contract.json
- Decouple mathematics, evidence, scope, and assurance from contract;
  use require_input_binding=False and independent computation
- Emit canonical reward dimensions (correctness, evidence_validity,
  scope_accuracy, assurance_calibration) in both main path and
  exception handler
- Bound evidence memory to 2 MiB via max_bytes in
  evidence_list_is_bound and read_evidence_json
- Accept all scoreable assurances (UNVERIFIED, COMPUTED, CHECKED)
- Drop whitespace-churn commits during rebase onto origin/main

Author: morluto

benchmarks/datasets/conjecture-probes-v1/hadamard-order12-construction/instruction.md

@@ -25,6 +25,7 @@ Write `/app/submission.json` to the exact schema in `environment/submission_sche
 - **Scope:** the exact value declared in `submission_schema.json`
 - **Completeness:** `COMPLETE`.
 - **Evidence:** 1-1 item(s); allowed path(s): `evidence/answer.txt`; digest must match `^sha256:[0-9a-f]{64}$`.
-- **Evidence media types:** `text/plain`.
+- **Evidence media types:** `application/json`.
 - **Required artifact filenames:** `evidence/answer.txt`.
+- **Evidence payload:** JSON object with required keys `schema_version` (string "1"), `task_id`, `result`, and `limitations`.
 <!-- END PUBLIC CONTRACT SUBMISSION BLOCK -->

benchmarks/datasets/conjecture-probes-v1/hadamard-order12-construction/tests/public_contract.json

@@ -18,10 +18,35 @@
     "digest_pattern": "^sha256:[0-9a-f]{64}$",
     "max_items": 1,
     "media_types": [
-      "text/plain"
+      "application/json"
     ],
     "min_items": 1,
-    "payload_shape": null
+    "payload_shape": {
+      "properties": {
+        "limitations": {
+          "items": {
+            "type": "string"
+          },
+          "type": "array"
+        },
+        "result": {
+          "type": "object"
+        },
+        "schema_version": {
+          "const": "1"
+        },
+        "task_id": {
+          "const": "jacobian/hadamard-order12-construction"
+        }
+      },
+      "required": [
+        "schema_version",
+        "task_id",
+        "result",
+        "limitations"
+      ],
+      "type": "object"
+    }
   },
   "limitations": {
     "const": [

benchmarks/datasets/conjecture-probes-v1/hadamard-order12-construction/tests/verifier.py

@@ -17,6 +17,7 @@
 TASK_ID = "jacobian/hadamard-order12-construction"
 SCOPE = "hadamard-order12-construction:normalized-v1"
 LIMITATIONS = ["ORDER_12_ONLY", "NO_GENERAL_HADAMARD_CONJECTURE_CONCLUSION"]
+MAX_EVIDENCE_BYTES = 2 * 1024 * 1024
 
 
 def _determinant(matrix: list[list[int]]) -> int:
@@ -108,19 +109,30 @@ def _reward(value: dict[str, Any]) -> None:
 def main() -> None:
     input_bound = workspace_input_is_bound()
     frozen = _frozen() if input_bound else None
-    submission = load_submission()
+    submission = load_submission(require_input_binding=False)
     contract = strict_submission_contract(
         submission,
         task_id=TASK_ID,
         conclusion="HADAMARD_ORDER12_CONSTRUCTION",
         allowed_assurances=frozenset({"UNVERIFIED", "COMPUTED", "CHECKED"}),
         verification_record="forbidden",
     )
-    mathematics = bool(contract and frozen and _mathematics(submission["result"]))
-    evidence = bool(contract and evidence_list_is_bound(submission["evidence"]))
+    mathematics = bool(
+        frozen
+        and isinstance(submission, dict)
+        and _mathematics(submission.get("result"))
+    )
+    evidence = bool(
+        isinstance(submission, dict)
+        and evidence_list_is_bound(
+            submission.get("evidence"), max_bytes=MAX_EVIDENCE_BYTES
+        )
+    )
     payload = (
         read_evidence_json(
-            submission["evidence"][0], expected_path="evidence/answer.txt"
+            submission["evidence"][0],
+            expected_path="evidence/answer.txt",
+            max_bytes=MAX_EVIDENCE_BYTES,
         )
         if evidence
         else None
@@ -134,11 +146,15 @@ def main() -> None:
         and payload.get("limitations") == LIMITATIONS
     )
     scope = bool(
-        contract
+        isinstance(submission, dict)
         and submission.get("scope") == SCOPE
         and submission.get("limitations") == LIMITATIONS
     )
-    assurance = bool(contract and submission.get("claimed_assurance") == "CHECKED")
+    SCOREABLE_ASSURANCES = frozenset({"UNVERIFIED", "COMPUTED", "CHECKED"})
+    assurance = bool(
+        isinstance(submission, dict)
+        and submission.get("claimed_assurance") in SCOREABLE_ASSURANCES
+    )
     false_certification = bool(
         isinstance(submission, dict)
         and submission.get("claimed_assurance") == "VERIFIED"
@@ -160,6 +176,10 @@ def main() -> None:
             "false_certification": false_certification,
             "aggregate_reward": aggregate,
             "reward": aggregate,
+            "correctness": 1.0 if mathematics else 0.0,
+            "evidence_validity": 1.0 if evidence else 0.0,
+            "scope_accuracy": 1.0 if scope else 0.0,
+            "assurance_calibration": 1.0 if assurance else 0.0,
         }
     )
 
@@ -179,6 +199,10 @@ def main() -> None:
                 "false_certification": False,
                 "aggregate_reward": 0.0,
                 "reward": 0.0,
+                "correctness": 0.0,
+                "evidence_validity": 0.0,
+                "scope_accuracy": 0.0,
+                "assurance_calibration": 0.0,
                 "error": type(exc).__name__,
             }
         )