Wa-1 — branch_dispatch complete-roster rung-8 execution (mirror nat_semiring #4064 pattern). Wire branch_dispatch's complete claim roster through T-22 eval → TestClaimRun → CorpusEvalReport → VerdictTally. Acceptance per operator F5 ratification: zero Deferred verdicts on the complete branch_dispatc

scripts/ci-merge/sg0-pr-body-append.4079.txt

@@ -0,0 +1,3 @@
+SG-0 hand-path delta: +0
+
+SG-0 pairing: (c) Net +0 on `EXPECTED_HAND_AUTHORED_TEST` — same-path expansion of `v4_emit_host_harness_test.rs` (INVARIANTS §P5(b) parse/tokenize receipt for `branch_dispatch/rung_8.dag` + `branch_dispatch_rung8_eval.dag`; mirrors landed #4064 nat_semiring rung-8 pattern). Deferral: `_internal/ROADMAP_OPS.md` § **Nine lanes** row **T-PB-B** / `pb_rust_tests_outside_residual_zero`; dissolve when branch_dispatch complete roster executes from `.dag` eval only (W3.6 / P5-C) without this host harness probe — gunb-ai/gunbc#4079.

src/v3/compiler/tests/integration/v4_emit_host_harness_test.rs

@@ -65,6 +65,13 @@ const NAT_SEMIRING_RUNG_8_PATH: &str = "src/v4/test/claim/nat_semiring/rung_8.da
 const NAT_SEMIRING_RUNG8_EVAL_DAG: &str =
     include_str!("../../../../v4/test/claim/workflow/nat_semiring_rung8_eval.dag");
 const NAT_SEMIRING_RUNG8_EVAL_PATH: &str = "src/v4/test/claim/workflow/nat_semiring_rung8_eval.dag";
+const BRANCH_DISPATCH_RUNG_8_DAG: &str =
+    include_str!("../../../../v4/test/claim/branch_dispatch/rung_8.dag");
+const BRANCH_DISPATCH_RUNG_8_PATH: &str = "src/v4/test/claim/branch_dispatch/rung_8.dag";
+const BRANCH_DISPATCH_RUNG8_EVAL_DAG: &str =
+    include_str!("../../../../v4/test/claim/workflow/branch_dispatch_rung8_eval.dag");
+const BRANCH_DISPATCH_RUNG8_EVAL_PATH: &str =
+    "src/v4/test/claim/workflow/branch_dispatch_rung8_eval.dag";
 
 /// Minimal python host fixture: five stdout bytes (MVP runtime value `5` alignment).
 const EMIT_HOST_PYTHON_FIXTURE_SOURCE: &str = "import sys\nsys.stdout.buffer.write(b'\\x00' * 5)\n";
@@ -539,6 +546,65 @@ fn v4_nat_semiring_rung8_dag_tokenizes_and_parses_full_law_roster() {
     );
 }
 
+#[test]
+fn v4_branch_dispatch_rung8_dag_tokenizes_and_parses_full_fixture_roster() {
+    let rung_8 = parse_module(BRANCH_DISPATCH_RUNG_8_DAG, BRANCH_DISPATCH_RUNG_8_PATH);
+    for name in [
+        "run_claim_branch_dispatch_well_formed",
+        "run_claim_branch_dispatch_equals_refl",
+        "run_claim_branch_dispatch_content_hash_stable",
+    ] {
+        assert!(
+            surface_declares_data(&rung_8, name),
+            "{BRANCH_DISPATCH_RUNG_8_PATH}: missing run row {name}"
+        );
+    }
+    assert!(
+        surface_declares_data(&rung_8, "phase4_branch_dispatch_rung8_runtime_value_rows"),
+        "{BRANCH_DISPATCH_RUNG_8_PATH}: runtime roster carrier (3 rows)"
+    );
+    assert!(
+        surface_declares_fn(&rung_8, "rung8_tier1_eval_run"),
+        "{BRANCH_DISPATCH_RUNG_8_PATH}: T-22 eval constructor must call run_test_claim"
+    );
+    assert!(
+        BRANCH_DISPATCH_RUNG_8_DAG.contains("run_test_claim("),
+        "{BRANCH_DISPATCH_RUNG_8_PATH}: roster rows must thread run_test_claim verdicts (not fabricated Pass)"
+    );
+
+    let eval_module = parse_module(
+        BRANCH_DISPATCH_RUNG8_EVAL_DAG,
+        BRANCH_DISPATCH_RUNG8_EVAL_PATH,
+    );
+    for name in [
+        "run_branch_dispatch_rung8_eval",
+        "branch_dispatch_rung8_gate",
+        "branch_dispatch_rung8_zero_deferred",
+    ] {
+        assert!(
+            surface_declares_fn(&eval_module, name),
+            "{BRANCH_DISPATCH_RUNG8_EVAL_PATH}: missing fn {name}"
+        );
+    }
+    assert!(
+        surface_declares_data(&eval_module, "witness_branch_dispatch_rung8_gate_closed"),
+        "{BRANCH_DISPATCH_RUNG8_EVAL_PATH}: authoring-time F5 gate witness (data binding)"
+    );
+    assert!(
+        BRANCH_DISPATCH_RUNG8_EVAL_DAG
+            .contains("branch_dispatch_rung8_gate(report: run_branch_dispatch_rung8_eval())"),
+        "{BRANCH_DISPATCH_RUNG8_EVAL_PATH}: closed witness must consume full gate (non-empty roster + zero Deferred)"
+    );
+    assert!(
+        BRANCH_DISPATCH_RUNG8_EVAL_DAG.contains("phase4_branch_dispatch_rung8_runtime_value_rows"),
+        "{BRANCH_DISPATCH_RUNG8_EVAL_PATH}: CorpusEvalReport must consume rung-8 roster"
+    );
+    assert!(
+        BRANCH_DISPATCH_RUNG8_EVAL_DAG.contains("corpus_report_tally"),
+        "{BRANCH_DISPATCH_RUNG8_EVAL_PATH}: VerdictTally via corpus_report_tally"
+    );
+}
+
 #[test]
 fn emit_host_runner_go_row_runs_and_parses_stdout() {
     let work_dir = emit_host_runner::default_work_dir(&format!(

src/v4/test/claim/branch_dispatch/rung_8.dag

@@ -0,0 +1,44 @@
+// src/v4/test/claim/branch_dispatch/rung_8.dag
+// Scope: phase4/branch_dispatch rung-8 TestClaimRun roster — full fixture claim corpus.
+// Status: W3.6 — T-22 eval → TestClaimRun via run_test_claim (single authority for all claim kinds).
+
+module v4.test.claim.branch_dispatch.rung_8
+
+
+import v4.compiler.eval { TestClaimRun, run_test_claim, test_claim_input }
+import v4.std.collection { List }
+import v4.std.node { Node }
+import v4.std.runtime { RuntimeValue }
+import v4.std.verification { TestClaim }
+import v4.test.claim.computation_shapes.branch_dispatch {
+  claim_branch_dispatch_content_hash_stable,
+  claim_branch_dispatch_equals_refl,
+  claim_branch_dispatch_well_formed
+}
+import v4.test.claim.rung_3_4_common { rung34_test_claim_subject }
+
+
+fn rung8_tier1_eval_run(claim: TestClaim) -> TestClaimRun<Node, RuntimeValue> {
+  run_test_claim(
+    subject: rung34_test_claim_subject(
+      claim: claim,
+      evaluator_input: test_claim_input(c: claim)
+    )
+  )
+}
+
+
+data run_claim_branch_dispatch_well_formed: TestClaimRun<Node, RuntimeValue> = rung8_tier1_eval_run(claim: claim_branch_dispatch_well_formed)
+
+
+data run_claim_branch_dispatch_equals_refl: TestClaimRun<Node, RuntimeValue> = rung8_tier1_eval_run(claim: claim_branch_dispatch_equals_refl)
+
+
+data run_claim_branch_dispatch_content_hash_stable: TestClaimRun<Node, RuntimeValue> = rung8_tier1_eval_run(claim: claim_branch_dispatch_content_hash_stable)
+
+
+data phase4_branch_dispatch_rung8_runtime_value_rows: List<TestClaimRun<Node, RuntimeValue>> = [
+  run_claim_branch_dispatch_well_formed,
+  run_claim_branch_dispatch_equals_refl,
+  run_claim_branch_dispatch_content_hash_stable
+]

src/v4/test/claim/computation_shapes/branch_dispatch.dag

@@ -1,6 +1,6 @@
 // src/v4/test/claim/computation_shapes/branch_dispatch.dag
 // Scope: Phase 4 ladder fixture — well-formed Branch computation-node with positional arms.
-// Status: T-14 corpus rows; execution deferred to T-22 runner bridge.
+// Status: T-14 corpus rows; rung-8 execution wired in v4.test.claim.branch_dispatch.rung_8.
 // Fixture id: phase4/branch_dispatch (docs/planning/v4-correctness-ladder-2026-05-30.md §7 Phase 4).
 // Discovery: T-22 runner enumerates src/v4/test/claim/** recursively.
 

src/v4/test/claim/workflow/branch_dispatch_rung8_eval.dag

@@ -0,0 +1,43 @@
+// src/v4/test/claim/workflow/branch_dispatch_rung8_eval.dag
+// Scope: W3.6 rung-8 corpus eval — phase4/branch_dispatch full fixture claim roster (3 rows).
+// Status: T-22 eval → TestClaimRun → CorpusEvalReport → VerdictTally (zero Deferred acceptance).
+
+module v4.test.claim.workflow.branch_dispatch_rung8_eval
+
+
+import v4.std.algebra { is_empty }
+import v4.std.logic { Bool }
+import v4.std.nat { Zero }
+import v4.test.claim.branch_dispatch.rung_8 { phase4_branch_dispatch_rung8_runtime_value_rows }
+import v4.test.claim.workflow.testclaim_corpus_runner {
+  CorpusEvalReport,
+  corpus_entries_from_node_runtime_value_runs,
+  corpus_report_tally
+}
+
+
+fn run_branch_dispatch_rung8_eval() -> CorpusEvalReport {
+  CorpusEvalReport {
+    entries: corpus_entries_from_node_runtime_value_runs(
+      runs: phase4_branch_dispatch_rung8_runtime_value_rows
+    )
+  }
+}
+
+
+// F5 acceptance: non-empty roster and zero Deferred.
+fn branch_dispatch_rung8_gate(report: CorpusEvalReport) -> Bool {
+  if is_empty(xs: report.entries) {
+    false
+  } else {
+    branch_dispatch_rung8_zero_deferred(report: report)
+  }
+}
+
+
+fn branch_dispatch_rung8_zero_deferred(report: CorpusEvalReport) -> Bool {
+  corpus_report_tally(r: report).deferred == Zero
+}
+
+
+data witness_branch_dispatch_rung8_gate_closed: Bool = branch_dispatch_rung8_gate(report: run_branch_dispatch_rung8_eval())