@@ -5,9 +5,9 @@ Machine-readable snapshot: `outputs/metrics_snapshot.json` regenerated with `mak
 > **Purpose:** One authoritative table of current pipeline metrics. If any doc disagrees
 > with this file, this file wins. Updated whenever benchmark/benchmark config changes.
 >
-> **Last updated:** 2026-08-22 (lab-result report schema validation; benchmark values unchanged)
+> **Last updated:** 2026-08-23 (pilot pre-registration lock enforcement; benchmark values unchanged)
 
-> **Current verification note (2026-08-22):** Phase AA AA6 exposes the AARG-
+> **Current verification note (2026-08-23):** Phase AA AA6 exposes the AARG-
 > reproducibility aggregate through a repeatable CLI/make workflow, while Phase
 > AB AB5 exposes the ABAG- claim-integrity aggregate, Phase AC AC3 exposes the
 > ACDG- aggregate, Phase Y Y5 exposes the YAG- aggregate, and Phase Z Z5
@@ -100,10 +100,16 @@ Machine-readable snapshot: `outputs/metrics_snapshot.json` regenerated with `mak
 > Phase AC AC3 exposes the ACDG-
 > aggregate disconfirming-evidence gate as a repeatable CLI/make workflow. It
 > has 18 focused gate tests plus 2 CLI integration tests. Full pytest
-> collection succeeds at 12,423 tests; the live count is enforced by the
+> collection succeeds at 12,424 tests; the live count is enforced by the
 > current-state alignment test. This artifact does not establish
 > biological validation or benchmark improvement.
 
+> **Pilot pre-registration lock note (2026-08-23):** The PRR- validator now
+> rejects unlocked records as invalid for experiment start. Unlocked drafts can
+> still be edited, but a valid pre-registration must freeze its selection
+> criteria and thresholds first. This is anti-cherry-picking infrastructure,
+> not assay validation or biological evidence.
+
 > **Canonical ERP generator note (2026-08-05):** `make generate-review-packet`
 > now exercises the V4 component-based ERP contract and emits a draft when no
 > artifact references are supplied. The legacy placeholder generator remains
 
@@ -1,6 +1,6 @@
 # Roadmap
 
-## Current state — 2026-08-22
+## Current state — 2026-08-23
 
 Phase AB is complete as of 2026-07-26. AB5 exposes the existing CSD-, RDR-,
 EGN-, and EHP- claim-integrity artifacts through an ABAG- aggregate, CLI
@@ -183,6 +183,12 @@ control-failure, raw-data, and input-blocker fields. This makes the report
 interoperable without treating report structure as assay validation, lab
 authentication, biological evidence, or recalibration authority.
 
+On 2026-08-23, pilot pre-registration validation began failing closed on
+unlocked records. Drafts remain editable, but only a locked PRR- record can
+validate as the pre-experiment contract. This enforces the existing
+pre-registration boundary; it does not validate assay procedures, biological
+claims, or partner readiness.
+
 This file is the current milestone authority. The older
 [`50_LOOP_PLAN.md`](50_LOOP_PLAN.md) is a historical execution record, not a
 live status page. Select the next bottleneck from
 
@@ -189,6 +189,10 @@ Required result context:
 
 Do not publish unsafe operational details.
 
+The machine validator treats `is_locked: true` as a validity requirement. An
+unlocked draft can be useful for internal editing, but it must not pass as a
+valid pre-registration for experiment start.
+
 ## Recalibration decision
 
 After result intake, run the pre-registered recalibration gate.
 
@@ -158,15 +158,22 @@ def validate_pilot_preregistration(
     if not record.frozen_at.strip():
         violations.append("frozen_at must not be empty")
 
-    # Rule 13: threshold values in [0, 1]
+    # Rule 13: lock required before a record can be valid
+    if not record.is_locked:
+        violations.append(
+            "is_locked must be True — selection criteria and thresholds must be "
+            "frozen before any experiment begins"
+        )
+
+    # Rule 14: threshold values in [0, 1]
     for t in record.score_thresholds:
         if not (0.0 <= t.threshold_value <= 1.0):
             violations.append(
                 f"score_thresholds[{t.score_name}].threshold_value must be in [0, 1], "
                 f"got {t.threshold_value}"
             )
 
-    # Rule 14: amendment_reasons vocabulary
+    # Rule 15: amendment_reasons vocabulary
     for reason in record.amendment_reasons:
         if reason not in VALID_AMENDMENT_REASONS:
             violations.append(
 
@@ -37,7 +37,7 @@ def _make_record(**kwargs) -> PilotPreregistration:
         negative_control="pbs_vehicle",
         outcome_metric="minimum_inhibitory_concentration",
         dry_lab_only_declaration=True,
-        is_locked=False,
+        is_locked=True,
     )
     defaults.update(kwargs)
     return PilotPreregistration(**defaults)
@@ -82,7 +82,14 @@ def test_default_amendment_count(self):
         assert r.amendment_count == 0
 
     def test_is_locked_default_false(self):
-        r = _make_record()
+        r = PilotPreregistration(
+            record_id="PRR-DEFAULT-001",
+            version="1.0.0",
+            frozen_at="2026-07-10T00:00:00Z",
+            pipeline_version="0.9.0",
+            git_sha="abc1234",
+            primary_hypothesis="draft",
+        )
         assert r.is_locked is False
 
 
@@ -112,6 +119,11 @@ def test_valid_record_passes(self):
         assert result.is_valid is True
         assert result.violations == []
 
+    def test_unlocked_record_is_not_valid_for_experiment_start(self):
+        result = validate_pilot_preregistration(_make_record(is_locked=False))
+        assert result.is_valid is False
+        assert any("is_locked" in violation for violation in result.violations)
+
     def test_returns_validation_result(self):
         result = validate_pilot_preregistration(_make_record())
         assert isinstance(result, PreregistrationValidationResult)