feat(governance): 新增 matched PlayBook 學習缺口

apps/api/tests/test_ai_agent_matched_playbook_learning_gap.py

@@ -0,0 +1,116 @@
+import copy
+import json
+
+import pytest
+
+from src.services.ai_agent_matched_playbook_learning_gap import (
+    load_latest_ai_agent_matched_playbook_learning_gap,
+)
+
+
+def _write_snapshot(tmp_path, payload):
+    path = tmp_path / "ai_agent_matched_playbook_learning_gap_2026-06-13.json"
+    path.write_text(json.dumps(payload), encoding="utf-8")
+    return path
+
+
+def test_load_latest_ai_agent_matched_playbook_learning_gap():
+    data = load_latest_ai_agent_matched_playbook_learning_gap()
+
+    assert data["schema_version"] == "ai_agent_matched_playbook_learning_gap_v1"
+    assert data["program_status"]["current_task_id"] == "P2-104"
+    assert data["program_status"]["next_task_id"] == "P2-105"
+    assert data["program_status"]["overall_completion_percent"] == 100
+    assert data["learning_gap_truth"]["matched_playbook_source_contract_ready"] is True
+    assert data["learning_gap_truth"]["approval_record_contract_ready"] is True
+    assert data["learning_gap_truth"]["learning_service_contract_ready"] is True
+    assert data["learning_gap_truth"]["runtime_execution_enabled"] is False
+    assert data["learning_gap_truth"]["live_db_read_enabled"] is False
+    assert data["learning_gap_truth"]["playbook_trust_write_enabled"] is False
+    assert data["learning_gap_truth"]["km_write_enabled"] is False
+    assert data["learning_gap_truth"]["timeline_write_enabled"] is False
+    assert data["learning_gap_truth"]["gateway_queue_write_enabled"] is False
+    assert data["learning_gap_truth"]["telegram_send_enabled"] is False
+    assert data["rollups"]["learning_checkpoint_count"] == 7
+    assert data["rollups"]["covered_by_test_count"] == 4
+    assert data["rollups"]["source_guarded_count"] == 2
+    assert data["rollups"]["blocked_until_live_evidence_count"] == 1
+    assert data["rollups"]["gap_item_count"] == 4
+    assert data["rollups"]["high_severity_gap_count"] == 2
+    assert data["rollups"]["test_contract_count"] == 5
+    assert data["rollups"]["owner_gate_count"] == 3
+    assert data["rollups"]["approval_required_gate_count"] == 3
+    assert data["rollups"]["live_db_read_count"] == 0
+    assert data["rollups"]["playbook_trust_write_count"] == 0
+    assert data["rollups"]["telegram_send_count"] == 0
+
+
+def test_rejects_live_db_read_enabled(tmp_path):
+    data = load_latest_ai_agent_matched_playbook_learning_gap()
+    bad = copy.deepcopy(data)
+    bad["learning_gap_truth"]["live_db_read_enabled"] = True
+    _write_snapshot(tmp_path, bad)
+
+    with pytest.raises(ValueError, match="live read/write/send/execution flags"):
+        load_latest_ai_agent_matched_playbook_learning_gap(tmp_path)
+
+
+def test_rejects_playbook_trust_write_count(tmp_path):
+    data = load_latest_ai_agent_matched_playbook_learning_gap()
+    bad = copy.deepcopy(data)
+    bad["learning_gap_truth"]["playbook_trust_write_count_24h"] = 1
+    bad["rollups"]["playbook_trust_write_count"] = 1
+    _write_snapshot(tmp_path, bad)
+
+    with pytest.raises(ValueError, match="live read/write/send/execution counts"):
+        load_latest_ai_agent_matched_playbook_learning_gap(tmp_path)
+
+
+def test_rejects_checkpoint_live_write(tmp_path):
+    data = load_latest_ai_agent_matched_playbook_learning_gap()
+    bad = copy.deepcopy(data)
+    bad["learning_checkpoints"][0]["writes_live_state"] = True
+    _write_snapshot(tmp_path, bad)
+
+    with pytest.raises(ValueError, match="writes_live_state"):
+        load_latest_ai_agent_matched_playbook_learning_gap(tmp_path)
+
+
+def test_rejects_checkpoint_without_evidence_ref(tmp_path):
+    data = load_latest_ai_agent_matched_playbook_learning_gap()
+    bad = copy.deepcopy(data)
+    bad["learning_checkpoints"][0]["evidence_ref"] = ""
+    _write_snapshot(tmp_path, bad)
+
+    with pytest.raises(ValueError, match="evidence_ref"):
+        load_latest_ai_agent_matched_playbook_learning_gap(tmp_path)
+
+
+def test_rejects_test_contract_live_execution(tmp_path):
+    data = load_latest_ai_agent_matched_playbook_learning_gap()
+    bad = copy.deepcopy(data)
+    bad["test_contracts"][0]["live_execution_required"] = True
+    _write_snapshot(tmp_path, bad)
+
+    with pytest.raises(ValueError, match="live_execution_required"):
+        load_latest_ai_agent_matched_playbook_learning_gap(tmp_path)
+
+
+def test_rejects_owner_gate_runtime_write(tmp_path):
+    data = load_latest_ai_agent_matched_playbook_learning_gap()
+    bad = copy.deepcopy(data)
+    bad["owner_gates"][0]["runtime_write_allowed"] = True
+    _write_snapshot(tmp_path, bad)
+
+    with pytest.raises(ValueError, match="runtime_write_allowed"):
+        load_latest_ai_agent_matched_playbook_learning_gap(tmp_path)
+
+
+def test_rejects_rollup_mismatch(tmp_path):
+    data = load_latest_ai_agent_matched_playbook_learning_gap()
+    bad = copy.deepcopy(data)
+    bad["rollups"]["learning_checkpoint_count"] = 999
+    _write_snapshot(tmp_path, bad)
+
+    with pytest.raises(ValueError, match="rollup counts"):
+        load_latest_ai_agent_matched_playbook_learning_gap(tmp_path)

apps/api/tests/test_ai_agent_matched_playbook_learning_gap_api.py

@@ -0,0 +1,37 @@
+from fastapi.testclient import TestClient
+
+from src.main import app
+
+
+def test_get_ai_agent_matched_playbook_learning_gap_api():
+    client = TestClient(app)
+    response = client.get("/api/v1/agents/agent-matched-playbook-learning-gap")
+
+    assert response.status_code == 200
+    data = response.json()
+    assert data["schema_version"] == "ai_agent_matched_playbook_learning_gap_v1"
+    assert data["program_status"]["current_task_id"] == "P2-104"
+    assert data["program_status"]["next_task_id"] == "P2-105"
+    assert data["program_status"]["overall_completion_percent"] == 100
+    assert data["learning_gap_truth"]["matched_playbook_source_contract_ready"] is True
+    assert data["learning_gap_truth"]["approval_record_contract_ready"] is True
+    assert data["learning_gap_truth"]["learning_service_contract_ready"] is True
+    assert data["learning_gap_truth"]["runtime_execution_enabled"] is False
+    assert data["learning_gap_truth"]["live_db_read_enabled"] is False
+    assert data["learning_gap_truth"]["playbook_trust_write_enabled"] is False
+    assert data["learning_gap_truth"]["km_write_enabled"] is False
+    assert data["learning_gap_truth"]["timeline_write_enabled"] is False
+    assert data["learning_gap_truth"]["gateway_queue_write_enabled"] is False
+    assert data["learning_gap_truth"]["telegram_send_enabled"] is False
+    assert data["rollups"]["learning_checkpoint_count"] == 7
+    assert data["rollups"]["covered_by_test_count"] == 4
+    assert data["rollups"]["source_guarded_count"] == 2
+    assert data["rollups"]["blocked_until_live_evidence_count"] == 1
+    assert data["rollups"]["gap_item_count"] == 4
+    assert data["rollups"]["high_severity_gap_count"] == 2
+    assert data["rollups"]["test_contract_count"] == 5
+    assert data["rollups"]["owner_gate_count"] == 3
+    assert data["rollups"]["approval_required_gate_count"] == 3
+    assert data["rollups"]["live_db_read_count"] == 0
+    assert data["rollups"]["playbook_trust_write_count"] == 0
+    assert data["rollups"]["telegram_send_count"] == 0