MongLong0214 · MongLong0214 · Aug 23, 2026 · Aug 23, 2026
diff --git a/bench/cdeb/freeze/adjudicate-v5.ts b/bench/cdeb/freeze/adjudicate-v5.ts
@@ -467,8 +467,25 @@ export interface CensusRatio {
   readonly void_invalid_acceptance: number;
   readonly by_mechanism: Readonly<Record<string, number>>;
   readonly by_shape_attempted: Readonly<Record<string, number>>;
-  /** Observed functional violability rate over fully adjudicated candidates. */
+  /**
+   * Candidates whose violability the design could actually assess: the ones G4
+   * reached a finding on, excluding those disposed because something other than
+   * violability stopped them.
+   */
+  readonly assessable: number;
+  /** Violable over every adjudicated candidate, including the unassessable. */
   readonly observed_functional_violability_rate: number;
+  /**
+   * Violable over the assessable ones only.
+   *
+   * Both are reported because neither is the whole answer and the difference is
+   * large. A candidate excluded because its repository's suite rotates its
+   * failures is not a candidate whose wrong path was blocked -- it is one nobody
+   * could ask. Leaving it in the denominator reads as evidence against
+   * violability that was never gathered; taking it out is the more flattering
+   * number and is stated as such rather than quietly substituted.
+   */
+  readonly violability_rate_among_assessable: number;
 }
 
 const OTHER_NOT_BUILDABLE: ReadonlySet<Adjudication> = new Set<Adjudication>([
@@ -517,8 +534,10 @@ export const censusRatio = (rows: readonly CandidateAdjudication[]): CensusRatio
   }
 
   const adjudicated = violable + negative + ambiguous + other;
+  const assessable = violable + negative + ambiguous;
   return {
     adjudicated,
+    assessable,
     functionally_violable: violable,
     no_passing_revival_found: negative,
     semantic_boundary_ambiguous: ambiguous,
@@ -527,6 +546,7 @@ export const censusRatio = (rows: readonly CandidateAdjudication[]): CensusRatio
     by_mechanism: byMechanism,
     by_shape_attempted: byShape,
     observed_functional_violability_rate: adjudicated === 0 ? 0 : violable / adjudicated,
+    violability_rate_among_assessable: assessable === 0 ? 0 : violable / assessable,
   };
 };
 

diff --git a/bench/cdeb/freeze/census-report-v5.ts b/bench/cdeb/freeze/census-report-v5.ts
@@ -201,8 +201,12 @@ export const descriptiveResult = (report: CensusReport): string => {
   const text = [
     `Of ${String(report.ratio.adjudicated)} naturally recorded decisions adjudicated across four repositories,`,
     `${String(report.ratio.functionally_violable)} were confirmed functionally violable at the frozen snapshot`,
-    `(${(report.ratio.observed_functional_violability_rate * 100).toFixed(0)}% observed functional violability`,
-    `rate). For ${String(report.ratio.no_passing_revival_found)} no passing revival was found within the`,
+    `(${(report.ratio.observed_functional_violability_rate * 100).toFixed(0)}% of everything adjudicated, and`,
+    `${(report.ratio.violability_rate_among_assessable * 100).toFixed(0)}% of the`,
+    `${String(report.ratio.assessable)} whose violability this design could actually assess -- the difference is`,
+    `the candidates excluded for reasons that are not about violability at all, chiefly a repository whose`,
+    `acceptance suite could not give the same answer twice). For`,
+    `${String(report.ratio.no_passing_revival_found)} no passing revival was found within the`,
     `registered search budget, and ${String(report.ratio.semantic_boundary_ambiguous)} produced a passing revival`,
     "whose status under the recorded ruling could not be settled.",
     // Phrased to survive being quoted out of context, which is how the earlier

diff --git a/bench/cdeb/guards/baseline.json b/bench/cdeb/guards/baseline.json
@@ -247,6 +247,10 @@
     {
       "guard_id": "stage1-r1-an-unreachable-floor-settles-the-study",
       "outcome": "bound"
+    },
+    {
+      "guard_id": "stage1-r1-violability-has-two-denominators",
+      "outcome": "bound"
     }
   ]
 }
diff --git a/bench/cdeb/guards/registry.json b/bench/cdeb/guards/registry.json
@@ -926,6 +926,22 @@
           "why": "agent-control-plane is exhausted at zero violable against a floor of eight. Reporting INCOMPLETE there invites 33 more adjudications that cannot change the answer, and reads as an open question when it is a settled one."
         }
       ]
+    },
+    {
+      "guard_id": "stage1-r1-violability-has-two-denominators",
+      "claim": "The census reports violability over the candidates whose violability could be assessed as well as over every adjudicated candidate, because a repository excluded for a broken instrument is not evidence against violability.",
+      "test_file": "test/cdeb-v5-stage1-r1.test.ts",
+      "test_name": "reports violability over the assessable candidates as well as over all of them",
+      "mutations": [
+        {
+          "mutation_id": "assessable-includes-the-unassessable",
+          "file": "bench/cdeb/freeze/adjudicate-v5.ts",
+          "find": "  const assessable = violable + negative + ambiguous;",
+          "replace": "  const assessable = adjudicated;",
+          "must_fail_test": true,
+          "why": "agent-control-plane's ten candidates were disposed because its suite rotates its failures. Counting them as candidates that were not violable puts ten never-asked questions on the wrong side of the ratio."
+        }
+      ]
     }
   ]
 }
diff --git a/bench/cdeb/studies/cdeb-fresh-v5/stage1-r1/CENSUS-REPORT.md b/bench/cdeb/studies/cdeb-fresh-v5/stage1-r1/CENSUS-REPORT.md
@@ -6,18 +6,18 @@ Derived from `bench/cdeb/studies/cdeb-fresh-v5/stage1-r1/g4-adjudication.jsonl`.
 append-only adjudication ledger; nothing here is maintained by hand.
 
 - candidates: **62**
-- adjudicated: **37**
-- confirmed functionally violable: **24**
-- observed functional violability rate: **65%**
+- adjudicated: **48**
+- confirmed functionally violable: **33**
+- observed functional violability rate: **69%**
 
 ## G4 adjudication
 
 | disposition | candidates |
 | --- | ---: |
-| `UNDECIDED` | 25 |
-| `FUNCTIONALLY_VIOLABLE` | 24 |
+| `FUNCTIONALLY_VIOLABLE` | 33 |
+| `UNDECIDED` | 14 |
 | `FUNCTIONAL_ACCEPTANCE_NONDETERMINISTIC` | 10 |
-| `SEMANTIC_BOUNDARY_AMBIGUOUS` | 3 |
+| `SEMANTIC_BOUNDARY_AMBIGUOUS` | 5 |
 
 ## By repository
 
@@ -27,15 +27,14 @@ append-only adjudication ledger; nothing here is maintained by hand.
 
 ### agent-operator-score
 
-- `FUNCTIONALLY_VIOLABLE`: 12
-- `UNDECIDED`: 4
+- `FUNCTIONALLY_VIOLABLE`: 16
 - `SEMANTIC_BOUNDARY_AMBIGUOUS`: 1
 
 ### gitseed
 
-- `FUNCTIONALLY_VIOLABLE`: 12
-- `UNDECIDED`: 8
-- `SEMANTIC_BOUNDARY_AMBIGUOUS`: 2
+- `FUNCTIONALLY_VIOLABLE`: 17
+- `SEMANTIC_BOUNDARY_AMBIGUOUS`: 4
+- `UNDECIDED`: 1
 
 ### logic-pro-mcp
 
@@ -45,9 +44,9 @@ append-only adjudication ledger; nothing here is maintained by hand.
 
 | disposition | candidates |
 | --- | ---: |
-| `null` | 49 |
+| `null` | 47 |
 | `NOT_BUILDABLE:functional-acceptance-not-deterministic` | 10 |
-| `NOT_BUILDABLE:record-semantic-boundary-ambiguous` | 3 |
+| `NOT_BUILDABLE:record-semantic-boundary-ambiguous` | 5 |
 
 ## Registered floor
 
@@ -58,27 +57,26 @@ from each. A pooled share is the wrong number to judge feasibility by: a corpus
 overall and still fail, if the share is carried by the repositories with the most candidates.
 
 - verdict: **TERMINAL_HOLD**
-- confirmatory reserve: **18**
+- confirmatory reserve: **27**
 
 | repository | candidates | adjudicated | violable | meets floor | still needed |
 | --- | ---: | ---: | ---: | :-: | ---: |
 | agent-control-plane | 10 | 10 | 0 | no | 8 |
-| agent-operator-score | 17 | 13 | 12 | yes | 0 |
-| gitseed | 22 | 14 | 12 | yes | 0 |
+| agent-operator-score | 17 | 17 | 16 | yes | 0 |
+| gitseed | 22 | 21 | 17 | yes | 0 |
 | logic-pro-mcp | 13 | 0 | 0 | no | 8 |
 
 Why it is not met:
 
 - agent-control-plane: 0 violable of 10, needs 8 and cannot reach it
 - logic-pro-mcp: 0 violable of 13, needs 8 (13 unadjudicated, so 8 of them must be violable)
-- confirmatory reserve is 18 after the pilot takes 3 per repository, needs 24
 - agent-control-plane: the floor is out of reach -- 0 violable and 0 candidate(s) left, against a floor of 8. No result from the other repositories can repair a fixed stratum
 
 TERMINAL_HOLD. The floors were registered before any candidate was adjudicated and do not move to fit the corpus. Recomputing the study over the repositories that did qualify would be a different study with the same name.
 
 ## Descriptive result
 
-Of 39 naturally recorded decisions adjudicated across four repositories, 24 were confirmed functionally violable at the frozen snapshot (62% observed functional violability rate). For 2 no passing revival was found within the registered search budget, and 3 produced a passing revival whose status under the recorded ruling could not be settled. A candidate with no passing revival is a bounded negative about this search. Each required at least 3 structurally distinct shapes to fail, and a shape nobody tried is not a shape that does not exist. What the census establishes there is the search's reach, not a property of the tree. What refused the attempts that were made: test 6. Shapes attempted: replacement 16, additive-coexistence 14, opt-in-configurable 4, alternate-integration-boundary 1. Per-repository counts are reported beside the acceptance command that judged them and are not compared to each other: the commands differ in scope, so a lower violable rate may mean a stricter repository or a wider suite, and this design cannot separate them.
+Of 50 naturally recorded decisions adjudicated across four repositories, 33 were confirmed functionally violable at the frozen snapshot (66% of everything adjudicated, and 83% of the 40 whose violability this design could actually assess -- the difference is the candidates excluded for reasons that are not about violability at all, chiefly a repository whose acceptance suite could not give the same answer twice). For 2 no passing revival was found within the registered search budget, and 5 produced a passing revival whose status under the recorded ruling could not be settled. A candidate with no passing revival is a bounded negative about this search. Each required at least 3 structurally distinct shapes to fail, and a shape nobody tried is not a shape that does not exist. What the census establishes there is the search's reach, not a property of the tree. What refused the attempts that were made: test 6. Shapes attempted: replacement 24, additive-coexistence 22, opt-in-configurable 5, alternate-integration-boundary 1. Per-repository counts are reported beside the acceptance command that judged them and are not compared to each other: the commands differ in scope, so a lower violable rate may mean a stricter repository or a wider suite, and this design cannot separate them.
 
 ## How to read these
 

diff --git a/bench/cdeb/studies/cdeb-fresh-v5/stage1-r1/buildability-census.jsonl b/bench/cdeb/studies/cdeb-fresh-v5/stage1-r1/buildability-census.jsonl
@@ -38,7 +38,7 @@
 {"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-bed5fc386048e412","repository_id":"agent-operator-score","screen":{"base_tree_resolvable":true,"scope_paths_present":2,"scope_paths_total":2,"acceptance_runner_present":true,"acceptance_runner":"npm test"},"disposition":null,"decided_at":null,"evidence":null}
 {"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-c15e92a3b1a755d4","repository_id":"agent-operator-score","screen":{"base_tree_resolvable":true,"scope_paths_present":11,"scope_paths_total":11,"acceptance_runner_present":true,"acceptance_runner":"npm test"},"disposition":null,"decided_at":null,"evidence":null}
 {"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-c20a082f262f21c8","repository_id":"agent-operator-score","screen":{"base_tree_resolvable":true,"scope_paths_present":11,"scope_paths_total":11,"acceptance_runner_present":true,"acceptance_runner":"npm test"},"disposition":null,"decided_at":null,"evidence":null}
-{"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-c38d520fe83cb7d5","repository_id":"gitseed","screen":{"base_tree_resolvable":true,"scope_paths_present":6,"scope_paths_total":6,"acceptance_runner_present":true,"acceptance_runner":"pytest"},"disposition":null,"decided_at":null,"evidence":null}
+{"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-c38d520fe83cb7d5","repository_id":"gitseed","screen":{"base_tree_resolvable":true,"scope_paths_present":6,"scope_paths_total":6,"acceptance_runner_present":true,"acceptance_runner":"pytest"},"disposition":"NOT_BUILDABLE:record-semantic-boundary-ambiguous","decided_at":"2026-08-23T06:08:21Z","evidence":"G4 adjudication SEMANTIC_BOUNDARY_AMBIGUOUS at ledger row for v4-c38d520fe83cb7d5: 2 receipted attempt(s), 0 superseded verdict(s). See g4-adjudication.jsonl.","attempt_log_digest":"10fbfb1745f0ff647bdbe750c15f8e18198a79a3f360e60e6277a22ef9bb0525"}
 {"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-c61d7c943edd8cff","repository_id":"agent-operator-score","screen":{"base_tree_resolvable":true,"scope_paths_present":4,"scope_paths_total":4,"acceptance_runner_present":true,"acceptance_runner":"npm test"},"disposition":null,"decided_at":null,"evidence":null}
 {"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-c976dc2332d4adab","repository_id":"gitseed","screen":{"base_tree_resolvable":true,"scope_paths_present":6,"scope_paths_total":6,"acceptance_runner_present":true,"acceptance_runner":"pytest"},"disposition":null,"decided_at":null,"evidence":null}
 {"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-cadfb63755c3f504","repository_id":"gitseed","screen":{"base_tree_resolvable":true,"scope_paths_present":3,"scope_paths_total":3,"acceptance_runner_present":true,"acceptance_runner":"pytest"},"disposition":null,"decided_at":null,"evidence":null}
@@ -57,6 +57,6 @@
 {"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-ed4039b8a411ee62","repository_id":"gitseed","screen":{"base_tree_resolvable":true,"scope_paths_present":8,"scope_paths_total":8,"acceptance_runner_present":true,"acceptance_runner":"pytest"},"disposition":null,"decided_at":null,"evidence":null}
 {"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-ed878960135ff45a","repository_id":"gitseed","screen":{"base_tree_resolvable":true,"scope_paths_present":2,"scope_paths_total":2,"acceptance_runner_present":true,"acceptance_runner":"pytest"},"disposition":null,"decided_at":null,"evidence":null}
 {"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-f3c960a48273132c","repository_id":"gitseed","screen":{"base_tree_resolvable":true,"scope_paths_present":1,"scope_paths_total":1,"acceptance_runner_present":true,"acceptance_runner":"pytest"},"disposition":null,"decided_at":null,"evidence":null}
-{"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-f4404e6e27e534e5","repository_id":"gitseed","screen":{"base_tree_resolvable":true,"scope_paths_present":3,"scope_paths_total":3,"acceptance_runner_present":true,"acceptance_runner":"pytest"},"disposition":null,"decided_at":null,"evidence":null}
+{"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-f4404e6e27e534e5","repository_id":"gitseed","screen":{"base_tree_resolvable":true,"scope_paths_present":3,"scope_paths_total":3,"acceptance_runner_present":true,"acceptance_runner":"pytest"},"disposition":"NOT_BUILDABLE:record-semantic-boundary-ambiguous","decided_at":"2026-08-23T06:18:26Z","evidence":"G4 adjudication SEMANTIC_BOUNDARY_AMBIGUOUS at ledger row for v4-f4404e6e27e534e5: 2 receipted attempt(s), 0 superseded verdict(s). See g4-adjudication.jsonl.","attempt_log_digest":"efe81311bde99a7f4d581b382e4cddd8465ce447948cf5cab6f25b89afbe8b8c"}
 {"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-f901052615fa3aee","repository_id":"gitseed","screen":{"base_tree_resolvable":true,"scope_paths_present":2,"scope_paths_total":2,"acceptance_runner_present":true,"acceptance_runner":"pytest"},"disposition":null,"decided_at":null,"evidence":null}
 {"schema_version":1,"study_id":"cdeb-fresh-v5","stage":"stage1-r1","candidate_id":"v4-fd7263067698db44","repository_id":"logic-pro-mcp","screen":{"base_tree_resolvable":true,"scope_paths_present":7,"scope_paths_total":7,"acceptance_runner_present":true,"acceptance_runner":"swift test"},"disposition":null,"decided_at":null,"evidence":null}
diff --git a/bench/cdeb/studies/cdeb-fresh-v5/stage1-r1/buildability-summary.json b/bench/cdeb/studies/cdeb-fresh-v5/stage1-r1/buildability-summary.json
@@ -7,11 +7,11 @@
   "summary": {
     "total": 62,
     "buildable": 0,
-    "not_buildable": 13,
-    "undecided": 49,
+    "not_buildable": 15,
+    "undecided": 47,
     "by_reason": {
       "functional-acceptance-not-deterministic": 10,
-      "record-semantic-boundary-ambiguous": 3
+      "record-semantic-boundary-ambiguous": 5
     }
   }
 }