Create expected_output.json

ProblemMap/Atlas/Fixes/official/demos/demo-f4-execution-closure/expected_output.json

@@ -0,0 +1,110 @@
+{
+  "demo_id": "demo_f4_execution_closure",
+  "demo_version": "v1",
+  "case_id": "f4_execution_case_001",
+  "expected_result_type": "route_first_repair_demo",
+  "minimum_success_contract": {
+    "primary_family": "F4",
+    "secondary_family": "F3",
+    "best_current_fit": "F4_N03 Pre-Readiness Execution Failure",
+    "broken_invariant": "execution_skeleton_closure_broken",
+    "fit_level": "node_level",
+    "confidence": "medium_or_higher",
+    "evidence_sufficiency": "sufficient_for_execution_first_cut"
+  },
+  "expected_output_contract": {
+    "final_answer_may_remain_placeholder": true,
+    "minimum_required_shift": "the repaired state must stop downstream execution from advancing on an invalid upstream condition",
+    "answer_quality_target": "structurally_valid_execution_path_restored",
+    "must_not_pretend": [
+      "full workflow robustness achieved",
+      "all downstream reasoning issues solved",
+      "final answer correctness already guaranteed in every variant"
+    ]
+  },
+  "expected_execution_contract": {
+    "must_expose": [
+      "retrieval_ready_gate_status",
+      "upstream_to_downstream_bridge_status",
+      "generation_execution_condition"
+    ],
+    "must_clarify": [
+      "what upstream condition was missing",
+      "why downstream execution should have been blocked",
+      "how closure becomes valid enough for the next stage"
+    ],
+    "execution_quality_target": "readiness_and_closure_restored_for_minimal_case"
+  },
+  "expected_route_explanation_contract": {
+    "why_primary_not_secondary_must_express": [
+      "the workflow advanced before a required upstream condition was satisfied",
+      "execution closure fails before continuity becomes the primary repair target"
+    ],
+    "must_not_reduce_case_to": [
+      "generic memory failure only",
+      "generic bad answer only",
+      "generic weak reasoning only"
+    ]
+  },
+  "expected_first_repair_move_contract": {
+    "must_include_repair_moves": [
+      "readiness_validation",
+      "ordering_validation",
+      "bridge_integrity_check",
+      "closure_path_trace",
+      "liveness_repair_if_needed"
+    ],
+    "must_not_start_with": [
+      "memory_strengthening",
+      "longer_chain_of_thought",
+      "generic_prompt_overhaul",
+      "retry_loops_without_gate_repair"
+    ]
+  },
+  "expected_before_after_contract": {
+    "before_workflow_state": "advanced_too_early",
+    "after_workflow_state": "blocked_until_ready_then_released",
+    "before_operator_position": "can_see_bad_output_but_not_yet_treat_it_as_a_closure_failure",
+    "after_operator_position": "can_identify_and_repair_the_broken_gate_and_bridge_path",
+    "required_change": "the repaired state must shift from premature downstream execution to a structurally valid gated path"
+  },
+  "expected_teaching_contract": {
+    "core_lesson": "If the workflow moved forward before closure existed, repair the skeleton first.",
+    "reader_should_learn": [
+      "not every workflow failure is a continuity failure",
+      "some failures are execution-first",
+      "correct routing changes the first repair move from continuity-oriented guessing to readiness and closure repair"
+    ]
+  },
+  "optional_wfgy_escalation_contract": {
+    "default_needed": false,
+    "allowed_when": [
+      "closure still fails after obvious gate repair",
+      "multiple bridge layers interact and local fixes create new instability",
+      "deeper liveness, retry, or cross-layer stress analysis is needed"
+    ],
+    "handoff_order": [
+      "atlas_route",
+      "first_execution_repair",
+      "wfgy_escalation_if_needed"
+    ]
+  },
+  "validation_notes": {
+    "json_role": "target_structure_for_demo_success",
+    "notebook_required_for_understanding": false,
+    "replay_mode_is_sufficient_for_teaching": true,
+    "community_variants_may_change_surface_wording": true,
+    "community_variants_must_preserve": [
+      "F4 as primary family",
+      "F3 as secondary family",
+      "readiness-first repair logic",
+      "the shift from premature execution to valid gated execution"
+    ]
+  },
+  "review_status": {
+    "schema_status": "ready",
+    "teaching_status": "ready",
+    "repair_contract_status": "ready",
+    "demo_alignment_status": "ready_for_notebook_and_fixture_checks"
+  }
+}