fix(sre): bind backup DR evidence into MCP

apps/api/src/jobs/agent99_controlled_dispatch_reconciler_job.py

@@ -503,6 +503,8 @@ async def _ensure_mcp_evidence_writeback(
         "rag_writeback_ack_id",
         "playbook_trust_writeback_ack_id",
     }
+    if "backup_health" in identity.route_id:
+        required.add("dr_scorecard_writeback_ack_id")
     safe_refs = sanitize_agent99_public_receipt_refs(
         receipt_refs,
         allowed_keys=required,
@@ -794,13 +796,6 @@ async def _finalize_learning(
             "telegram_lifecycle_receipt_id",
             lambda: _ensure_telegram_receipt(identity, mode=mode),
         ),
-        (
-            "mcp_evidence_writeback_ack_id",
-            lambda: _ensure_mcp_evidence_writeback(
-                identity,
-                receipt_refs=receipt_refs,
-            ),
-        ),
     ]
     if "backup_health" in identity.route_id:
         assets.append((
@@ -810,6 +805,18 @@ async def _finalize_learning(
                 evidence_refs=evidence_refs,
             ),
         ))
+    # MCP is the final evidence binding: for backup routes it must include the
+    # already-persisted DR scorecard acknowledgement as well as Telegram,
+    # KM/RAG and PlayBook.  Writing MCP first would create a durable but
+    # incomplete evidence graph that could never prove the DR asset belonged
+    # to this same run.
+    assets.append((
+        "mcp_evidence_writeback_ack_id",
+        lambda: _ensure_mcp_evidence_writeback(
+            identity,
+            receipt_refs=receipt_refs,
+        ),
+    ))
     for key, writer in assets:
         if not await ensure_checkpointed(key, writer):
             return {

apps/api/tests/test_agent99_controlled_dispatch_reconciler_job.py

@@ -147,6 +147,28 @@ async def test_agent99_mcp_ack_is_same_run_durable_and_idempotent(
     assert second == expected
 
 
+@pytest.mark.asyncio
+async def test_backup_mcp_ack_requires_dr_scorecard_binding() -> None:
+    identity = build_agent99_dispatch_identity(
+        project_id="awoooi",
+        incident_id="INC-20260719-BACKUP-MCP",
+        source_fingerprint="backup-mcp-fingerprint",
+        route_id="agent99:backup_health:BackupCheck",
+    )
+
+    receipt = await job._ensure_mcp_evidence_writeback(
+        identity,
+        receipt_refs={
+            "telegram_lifecycle_receipt_id": "telegram:1",
+            "km_writeback_ack_id": "km:1",
+            "rag_writeback_ack_id": "rag:1",
+            "playbook_trust_writeback_ack_id": "playbook:1",
+        },
+    )
+
+    assert receipt is None
+
+
 @pytest.mark.asyncio
 async def test_backup_telegram_ack_is_no_write_lifecycle_not_apply(
     monkeypatch,
@@ -445,30 +467,40 @@ async def test_learning_checkpoint_resumes_without_replaying_assets(
         "mcp": 0,
         "dr": 0,
     }
+    effect_order: list[str] = []
     terminal_calls = 0
 
     async def write_playbook(*_args, **_kwargs):  # type: ignore[no-untyped-def]
         effects["playbook"] += 1
+        effect_order.append("playbook")
         return "playbooks:PB-A99:version:1"
 
     async def write_km(*_args, **_kwargs):  # type: ignore[no-untyped-def]
         effects["km"] += 1
+        effect_order.append("km")
         return "knowledge_entries:km-1:embedding_verified"
 
     async def write_telegram(*_args, **_kwargs):  # type: ignore[no-untyped-def]
         effects["telegram"] += 1
+        effect_order.append("telegram")
         return "telegram_outbound:1:durable_ack"
 
     async def write_rag(*_args, **_kwargs):  # type: ignore[no-untyped-def]
         effects["rag"] += 1
+        effect_order.append("rag")
         return "knowledge_entries:km-1:rag_embedding_verified"
 
     async def write_mcp(*_args, **_kwargs):  # type: ignore[no-untyped-def]
         effects["mcp"] += 1
+        effect_order.append("mcp")
+        assert _kwargs["receipt_refs"]["dr_scorecard_writeback_ack_id"] == (
+            "knowledge_entries:dr-1:dr_scorecard"
+        )
         return "awooop_mcp_gateway_audit:call-1:verified"
 
     async def write_dr(*_args, **_kwargs):  # type: ignore[no-untyped-def]
         effects["dr"] += 1
+        effect_order.append("dr")
         return "knowledge_entries:dr-1:dr_scorecard"
 
     async def terminal(**_kwargs):  # type: ignore[no-untyped-def]
@@ -514,6 +546,7 @@ async def test_learning_checkpoint_resumes_without_replaying_assets(
         "mcp": 1,
         "dr": 1,
     }
+    assert effect_order == ["playbook", "km", "rag", "telegram", "dr", "mcp"]
     assert terminal_calls == 2