diff --git a/apps/api/tests/test_agent99_controlled_dispatch_p1.py b/apps/api/tests/test_agent99_controlled_dispatch_p1.py
index 3b9104734..0a09da8aa 100644
--- a/apps/api/tests/test_agent99_controlled_dispatch_p1.py
+++ b/apps/api/tests/test_agent99_controlled_dispatch_p1.py
@@ -41,6 +41,16 @@ def _evidence_refs() -> dict[str, str]:
     }
 
 
+def _learning_receipt_refs() -> dict[str, str]:
+    return {
+        "telegram_lifecycle_receipt_id": "telegram:1",
+        "km_writeback_ack_id": "knowledge:1",
+        "rag_writeback_ack_id": "rag:1",
+        "mcp_evidence_writeback_ack_id": "mcp:1",
+        "playbook_trust_writeback_ack_id": "playbook:1",
+    }
+
+
 def _outcome(identity, *, transport_ok: bool = True) -> dict:
     return {
         "identity": identity.public_dict(),
@@ -580,11 +590,7 @@ async def test_learning_writeback_is_the_only_same_run_terminal_step(
     )
     incomplete = await ledger.record_learning_writeback(
         identity=identity,
-        receipt_refs={
-            "telegram_lifecycle_receipt_id": "telegram:fake",
-            "km_writeback_ack_id": "knowledge:fake",
-            "playbook_trust_writeback_ack_id": "playbook:fake",
-        },
+        receipt_refs=_learning_receipt_refs(),
     )
     assert incomplete["status"] == (
         "learning_writeback_checkpoint_incomplete_fail_closed"
@@ -592,11 +598,7 @@ async def test_learning_writeback_is_the_only_same_run_terminal_step(
     assert incomplete["runtime_closure_verified"] is False
     assert incomplete_db.call == 1
 
-    checkpoint_refs = {
-        "telegram_lifecycle_receipt_id": "telegram-1",
-        "km_writeback_ack_id": "km-1",
-        "playbook_trust_writeback_ack_id": "playbook-1",
-    }
+    checkpoint_refs = _learning_receipt_refs()
     verifier_passed["learning_writeback"]["receipt_refs"] = checkpoint_refs
 
     class LearningDB:
diff --git a/apps/api/tests/test_sre_k3s_controlled_automation_work_items_api.py b/apps/api/tests/test_sre_k3s_controlled_automation_work_items_api.py
index 0082df2b8..c5a2607db 100644
--- a/apps/api/tests/test_sre_k3s_controlled_automation_work_items_api.py
+++ b/apps/api/tests/test_sre_k3s_controlled_automation_work_items_api.py
@@ -116,11 +116,11 @@ def test_loader_returns_fixed_architecture_provider_order_and_agent99_bridge() -
         "total_items": 18,
         "by_priority": {"P0": 16, "P1": 2},
         "by_status": {
-            "source_implemented_runtime_pending": 11,
-            "in_progress": 3,
+            "source_implemented_runtime_pending": 12,
+            "in_progress": 2,
             "planned": 4,
         },
-        "source_implemented_items": 11,
+        "source_implemented_items": 12,
         "runtime_closed_items": 0,
         "program_completion_percent": 0,
         "asset_coverage_status": "partial",
@@ -141,6 +141,12 @@ def test_ledger_links_confirmed_runtime_gaps_without_false_closure() -> None:
         items["AIA-SRE-013"]["runtime_gaps"]
     )
     assert "host99 Agent99" in " ".join(items["AIA-SRE-017"]["runtime_gaps"])
+    assert items["AIA-SRE-009"]["status"] == (
+        "source_implemented_runtime_pending"
+    )
+    assert "arbitrary commands" in " ".join(
+        items["AIA-SRE-009"]["confirmed_truth"]
+    )
     assert items["AIA-SRE-011"]["status"] == (
         "source_implemented_runtime_pending"
     )
@@ -177,7 +183,7 @@ def test_projection_keeps_program_asset_and_runtime_truth_separate() -> None:
     payload = load_sre_k3s_controlled_automation_work_items()
     projection = build_sre_k3s_program_projection(payload)
 
-    assert projection["rollups"]["source_implemented_items"] == 11
+    assert projection["rollups"]["source_implemented_items"] == 12
     assert projection["rollups"]["runtime_closed_items"] == 0
     assert projection["rollups"]["program_completion_percent"] == 0
     assert projection["domain_count"] == 8
diff --git a/docs/operations/sre-k3s-controlled-automation-work-items.snapshot.json b/docs/operations/sre-k3s-controlled-automation-work-items.snapshot.json
index ec7d8516b..3cf5aa236 100644
--- a/docs/operations/sre-k3s-controlled-automation-work-items.snapshot.json
+++ b/docs/operations/sre-k3s-controlled-automation-work-items.snapshot.json
@@ -512,19 +512,33 @@
       "title": "Windows/VMware Agent99 single-executor closure",
       "owner_lane": "Agent99",
       "risk": "high",
-      "status": "in_progress",
+      "status": "source_implemented_runtime_pending",
       "dependencies": [
         "AIA-SRE-006"
       ],
       "source_refs": [
         "apps/api/src/services/agent99_controlled_dispatch_ledger.py",
-        "apps/api/src/services/agent99_same_run_reconcile.py"
+        "apps/api/src/services/agent99_same_run_reconcile.py",
+        "apps/api/src/services/agent99_sre_bridge.py",
+        "apps/api/src/services/agent99_completion_callback.py",
+        "apps/api/src/services/agent99_outcome_ingestion.py",
+        "apps/api/src/services/agent99_public_receipts.py",
+        "apps/api/src/jobs/agent99_controlled_dispatch_reconciler_job.py"
       ],
       "executor": "Agent99",
       "verifier": "agent99_independent_runtime_verifier",
       "rollback": "allowlisted Status/no-write reconciliation and bounded generation retry",
       "exit_condition": "Windows/VMware runs use durable identity, callback, verifier and learning receipt",
-      "next_action": "reconcile remaining Agent99 enterprise work ledger with this program"
+      "confirmed_truth": [
+        "Windows/VMware and control-plane recovery dispatches require a typed same-run identity and Agent99 executor binding",
+        "unknown assets, arbitrary commands and cross-domain fallback fail closed before runtime authorization",
+        "authenticated completion callback persists outcome and independent verifier evidence before learning writeback",
+        "terminal closure requires checkpointed Telegram, KM, RAG, MCP and PlayBook acknowledgements on the same run"
+      ],
+      "runtime_gaps": [
+        "no production Windows/VMware same-run dispatch, authenticated callback, independent verifier and complete learning closure receipt is recorded for this exact source revision"
+      ],
+      "next_action": "deploy one exact source revision through the authorized Agent99 release lane, replay one allowlisted Windows/VMware run plus one cross-domain denial, then read back the same-run callback, verifier, Telegram and learning receipts; do not use arbitrary commands or fallback hosts"
     },
     {
       "id": "AIA-SRE-010",
@@ -847,11 +861,11 @@
       "P1": 2
     },
     "by_status": {
-      "source_implemented_runtime_pending": 11,
-      "in_progress": 3,
+      "source_implemented_runtime_pending": 12,
+      "in_progress": 2,
       "planned": 4
     },
-    "source_implemented_items": 11,
+    "source_implemented_items": 12,
     "runtime_closed_items": 0,
     "program_completion_percent": 0,
     "asset_coverage_status": "partial",