ARCH: замкнуть metadata-scoped clarification recovery loops

2026-04-23 13:42:14 +03:00
parent c96c9bab86
commit cddd6667fb
20 changed files with 993 additions and 43 deletions
@@ -229,4 +229,26 @@ describe("assistant MCP discovery data need graph", () => {
    expect(result.reason_codes).toContain("data_need_graph_open_scope_total_without_subject");
    expect(result.reason_codes).toContain("data_need_graph_all_time_scope_hint");
  });
+
+  it("treats metadata-scoped movement evidence as subjectless and asks only for organization plus period", () => {
+    const result = buildAssistantMcpDiscoveryDataNeedGraph({
+      semanticDataNeed: "movement evidence",
+      rawUtterance: "РїРѕ РґРІРёР¶РµРЅРёСЏРј",
+      turnMeaning: {
+        asked_domain_family: "movements",
+        asked_action_family: "list_movements",
+        metadata_scope_hint: "РќР”РЎ",
+        subject_resolution_optional: true,
+        unsupported_but_understood_family: "movement_evidence"
+      }
+    });
+
+    expect(result.business_fact_family).toBe("movement_evidence");
+    expect(result.subject_candidates).toEqual([]);
+    expect(result.metadata_scope_hint).toBe("РќР”РЎ");
+    expect(result.subject_resolution_optional).toBe(true);
+    expect(result.clarification_gaps).toEqual(["organization", "period"]);
+    expect(result.decomposition_candidates).toEqual(["fetch_scoped_movements", "probe_coverage"]);
+    expect(result.reason_codes).toContain("data_need_graph_metadata_scoped_open_lane_without_subject");
+  });
 });
@@ -823,4 +823,86 @@ describe("assistant MCP discovery planner", () => {
    expect(result.catalog_review.review_status).toBe("catalog_compatible");
    expect(result.reason_codes).toContain("planner_selected_open_scope_value_flow_total_from_data_need_graph");
  });
+
+  it("keeps metadata-scoped movement evidence in clarification instead of forcing entity resolution", () => {
+    const result = planAssistantMcpDiscovery({
+      dataNeedGraph: {
+        schema_version: "assistant_data_need_graph_v1",
+        policy_owner: "assistantMcpDiscoveryDataNeedGraph",
+        subject_candidates: [],
+        metadata_scope_hint: "РќР”РЎ",
+        subject_resolution_optional: true,
+        business_fact_family: "movement_evidence",
+        action_family: "list_movements",
+        aggregation_need: null,
+        time_scope_need: "period_required",
+        comparison_need: null,
+        ranking_need: null,
+        proof_expectation: "clarification_required",
+        clarification_gaps: ["organization", "period"],
+        decomposition_candidates: ["fetch_scoped_movements", "probe_coverage"],
+        forbidden_overclaim_flags: ["no_raw_model_claims", "no_unchecked_fact_totals"],
+        reason_codes: [
+          "data_need_graph_built",
+          "data_need_graph_metadata_scoped_open_lane_without_subject"
+        ]
+      },
+      turnMeaning: {
+        asked_domain_family: "movements",
+        asked_action_family: "list_movements",
+        metadata_scope_hint: "РќР”РЎ",
+        subject_resolution_optional: true,
+        unsupported_but_understood_family: "movement_evidence"
+      }
+    });
+
+    expect(result.planner_status).toBe("needs_clarification");
+    expect(result.selected_chain_id).toBe("movement_evidence");
+    expect(result.proposed_primitives).toEqual(["query_movements", "probe_coverage"]);
+    expect(result.required_axes).toEqual(["metadata_scope", "organization", "coverage_target"]);
+    expect(result.reason_codes).toContain("planner_selected_metadata_scoped_movement_from_data_need_graph");
+    expect(result.selected_chain_id).not.toBe("entity_resolution");
+  });
+
+  it("keeps metadata-scoped movement evidence execution-ready once organization and period are known", () => {
+    const result = planAssistantMcpDiscovery({
+      dataNeedGraph: {
+        schema_version: "assistant_data_need_graph_v1",
+        policy_owner: "assistantMcpDiscoveryDataNeedGraph",
+        subject_candidates: [],
+        metadata_scope_hint: "РќР”РЎ",
+        subject_resolution_optional: true,
+        business_fact_family: "movement_evidence",
+        action_family: "list_movements",
+        aggregation_need: null,
+        time_scope_need: "explicit_period",
+        comparison_need: null,
+        ranking_need: null,
+        proof_expectation: "coverage_checked_fact",
+        clarification_gaps: [],
+        decomposition_candidates: ["fetch_scoped_movements", "probe_coverage"],
+        forbidden_overclaim_flags: ["no_raw_model_claims", "no_unchecked_fact_totals"],
+        reason_codes: [
+          "data_need_graph_built",
+          "data_need_graph_metadata_scoped_open_lane_without_subject"
+        ]
+      },
+      turnMeaning: {
+        asked_domain_family: "movements",
+        asked_action_family: "list_movements",
+        metadata_scope_hint: "РќР”РЎ",
+        subject_resolution_optional: true,
+        explicit_organization_scope: "РћРћРћ РђР»СЊС‚РµСЂРЅР°С‚РёРІР° РџР»СЋСЃ",
+        explicit_date_scope: "2020",
+        unsupported_but_understood_family: "movement_evidence"
+      }
+    });
+
+    expect(result.planner_status).toBe("ready_for_execution");
+    expect(result.selected_chain_id).toBe("movement_evidence");
+    expect(result.proposed_primitives).toEqual(["query_movements", "probe_coverage"]);
+    expect(result.required_axes).toEqual(["organization", "period", "metadata_scope", "coverage_target"]);
+    expect(result.catalog_review.review_status).toBe("catalog_compatible");
+    expect(result.reason_codes).toContain("planner_selected_metadata_scoped_movement_from_data_need_graph");
+  });
 });
@@ -429,4 +429,44 @@ describe("assistant MCP discovery runtime bridge", () => {
      "\u043a\u043e\u043d\u0442\u0440\u0430\u0433\u0435\u043d\u0442\u0430"
    );
  });
+  it("persists metadata scope and subject-optional flags in the resumable loop state", async () => {
+    const result = await runAssistantMcpDiscoveryRuntimeBridge({
+      dataNeedGraph: {
+        schema_version: "assistant_data_need_graph_v1",
+        policy_owner: "assistantMcpDiscoveryDataNeedGraph",
+        subject_candidates: [],
+        metadata_scope_hint: "\u041d\u0414\u0421",
+        subject_resolution_optional: true,
+        business_fact_family: "movement_evidence",
+        action_family: "list_movements",
+        aggregation_need: null,
+        time_scope_need: "period_required",
+        comparison_need: null,
+        ranking_need: null,
+        proof_expectation: "clarification_required",
+        clarification_gaps: ["organization", "period"],
+        decomposition_candidates: ["fetch_scoped_movements", "probe_coverage"],
+        forbidden_overclaim_flags: ["no_raw_model_claims", "no_unchecked_fact_totals"],
+        reason_codes: ["data_need_graph_built", "data_need_graph_metadata_scoped_open_lane_without_subject"]
+      },
+      turnMeaning: {
+        asked_domain_family: "movements",
+        asked_action_family: "list_movements",
+        metadata_scope_hint: "\u041d\u0414\u0421",
+        subject_resolution_optional: true,
+        unsupported_but_understood_family: "movement_evidence"
+      },
+      deps: buildDeps([])
+    });
+
+    expect(result.bridge_status).toBe("needs_clarification");
+    expect(result.loop_state).toMatchObject({
+      loop_status: "awaiting_clarification",
+      selected_chain_id: "movement_evidence",
+      metadata_scope_hint: "\u041d\u0414\u0421",
+      subject_resolution_optional: true
+    });
+    expect(result.loop_state.pending_axes).toEqual(["organization", "period"]);
+    expect(result.loop_state.explicit_entity_candidates).toEqual([]);
+  });
 });
@@ -1877,4 +1877,144 @@ describe("assistant MCP discovery turn input adapter", () => {
    );
    expect(result.reason_codes).not.toContain("mcp_discovery_not_applicable_for_supported_exact_turn");
  });
+
+  it.skip("keeps metadata-born movement lane subjectless and asks for organization plus period", () => {
+    const result = buildAssistantMcpDiscoveryTurnInput({
+      userMessage: "РїРѕ РґРІРёР¶РµРЅРёСЏРј",
+      followupContext: {
+        previous_discovery_loop_status: "awaiting_clarification",
+        previous_discovery_loop_selected_chain_id: "metadata_lane_clarification",
+        previous_discovery_loop_pending_axes: ["lane_family_choice"],
+        previous_discovery_loop_asked_domain_family: "metadata",
+        previous_discovery_loop_asked_action_family: "resolve_next_lane",
+        previous_discovery_loop_unsupported_family: "metadata_lane_choice_clarification",
+        previous_discovery_entity_candidates: ["РќР”РЎ"],
+        previous_discovery_metadata_ambiguity_detected: true,
+        previous_discovery_metadata_ambiguity_entity_sets: ["Р”РѕРєСѓРјРµРЅС‚", "Р РµРіРёСЃС‚СЂРќР°РєРѕРїР»РµРЅРёСЏ"]
+      }
+    });
+
+    expect(result.adapter_status).toBe("ready");
+    expect(result.should_run_discovery).toBe(true);
+    expect(result.semantic_data_need).toBe("movement evidence");
+    expect(result.turn_meaning_ref).toMatchObject({
+      asked_domain_family: "movements",
+      asked_action_family: "list_movements",
+      metadata_scope_hint: "РќР”РЎ",
+      subject_resolution_optional: true,
+      unsupported_but_understood_family: "movement_evidence",
+      stale_replay_forbidden: true
+    });
+    expect(result.turn_meaning_ref?.explicit_entity_candidates).toBeUndefined();
+    expect(result.data_need_graph?.clarification_gaps).toEqual(["organization", "period"]);
+    expect(result.reason_codes).toContain("mcp_discovery_metadata_ambiguity_resolved_to_movement_lane");
+    expect(result.reason_codes).toContain("mcp_discovery_metadata_scoped_lane_without_subject");
+  });
+
+  it.skip("keeps metadata scope through organization-only clarification and leaves only period pending", () => {
+    const orgName = "РћРћРћ РђР»СЊС‚РµСЂРЅР°С‚РёРІР° РџР»СЋСЃ";
+    const result = buildAssistantMcpDiscoveryTurnInput({
+      userMessage: "РїРѕ РћРћРћ РђР»СЊС‚РµСЂРЅР°С‚РёРІР° РџР»СЋСЃ",
+      predecomposeContract: {
+        entities: { organization: orgName }
+      },
+      followupContext: {
+        previous_discovery_loop_status: "awaiting_clarification",
+        previous_discovery_loop_selected_chain_id: "movement_evidence",
+        previous_discovery_loop_pending_axes: ["organization", "period"],
+        previous_discovery_loop_asked_domain_family: "movements",
+        previous_discovery_loop_asked_action_family: "list_movements",
+        previous_discovery_loop_unsupported_family: "movement_evidence",
+        previous_discovery_entity_candidates: ["РќР”РЎ"]
+      }
+    });
+
+    expect(result.adapter_status).toBe("ready");
+    expect(result.should_run_discovery).toBe(true);
+    expect(result.semantic_data_need).toBe("movement evidence");
+    expect(result.turn_meaning_ref).toMatchObject({
+      asked_domain_family: "movements",
+      asked_action_family: "list_movements",
+      metadata_scope_hint: "РќР”РЎ",
+      subject_resolution_optional: true,
+      explicit_organization_scope: orgName,
+      unsupported_but_understood_family: "movement_evidence",
+      stale_replay_forbidden: true
+    });
+    expect(result.turn_meaning_ref?.explicit_entity_candidates).toBeUndefined();
+    expect(result.data_need_graph?.clarification_gaps).toEqual(["period"]);
+  });
+  it("keeps metadata-born movement lane subjectless and asks for organization plus period (utf8-safe)", () => {
+    const result = buildAssistantMcpDiscoveryTurnInput({
+      userMessage: "\u043f\u043e \u0434\u0432\u0438\u0436\u0435\u043d\u0438\u044f\u043c",
+      followupContext: {
+        previous_discovery_loop_status: "awaiting_clarification",
+        previous_discovery_loop_selected_chain_id: "metadata_lane_clarification",
+        previous_discovery_loop_pending_axes: ["lane_family_choice"],
+        previous_discovery_loop_asked_domain_family: "metadata",
+        previous_discovery_loop_asked_action_family: "resolve_next_lane",
+        previous_discovery_loop_unsupported_family: "metadata_lane_choice_clarification",
+        previous_discovery_entity_candidates: ["\u041d\u0414\u0421"],
+        previous_discovery_metadata_ambiguity_detected: true,
+        previous_discovery_metadata_ambiguity_entity_sets: [
+          "\u0414\u043e\u043a\u0443\u043c\u0435\u043d\u0442",
+          "\u0420\u0435\u0433\u0438\u0441\u0442\u0440\u041d\u0430\u043a\u043e\u043f\u043b\u0435\u043d\u0438\u044f"
+        ]
+      }
+    });
+
+    expect(result.adapter_status).toBe("ready");
+    expect(result.should_run_discovery).toBe(true);
+    expect(result.semantic_data_need).toBe("movement evidence");
+    expect(result.turn_meaning_ref).toMatchObject({
+      asked_domain_family: "movements",
+      asked_action_family: "list_movements",
+      metadata_scope_hint: "\u041d\u0414\u0421",
+      subject_resolution_optional: true,
+      unsupported_but_understood_family: "movement_evidence",
+      stale_replay_forbidden: true
+    });
+    expect(result.turn_meaning_ref?.explicit_entity_candidates).toBeUndefined();
+    expect(result.data_need_graph?.clarification_gaps).toEqual(["organization", "period"]);
+    expect(result.reason_codes).toContain("mcp_discovery_metadata_ambiguity_resolved_to_movement_lane");
+    expect(result.reason_codes).toContain("mcp_discovery_metadata_scoped_lane_without_subject");
+  });
+
+  it("keeps metadata scope through organization-only clarification and leaves only period pending (utf8-safe)", () => {
+    const orgName =
+      "\u041e\u041e\u041e \u0410\u043b\u044c\u0442\u0435\u0440\u043d\u0430\u0442\u0438\u0432\u0430 \u041f\u043b\u044e\u0441";
+    const result = buildAssistantMcpDiscoveryTurnInput({
+      userMessage:
+        "\u043f\u043e \u041e\u041e\u041e \u0410\u043b\u044c\u0442\u0435\u0440\u043d\u0430\u0442\u0438\u0432\u0430 \u041f\u043b\u044e\u0441",
+      predecomposeContract: {
+        entities: { organization: orgName }
+      },
+      followupContext: {
+        previous_discovery_loop_status: "awaiting_clarification",
+        previous_discovery_loop_selected_chain_id: "movement_evidence",
+        previous_discovery_loop_pending_axes: ["organization", "period"],
+        previous_discovery_loop_asked_domain_family: "movements",
+        previous_discovery_loop_asked_action_family: "list_movements",
+        previous_discovery_loop_unsupported_family: "movement_evidence",
+        previous_discovery_loop_metadata_scope_hint: "\u041d\u0414\u0421",
+        previous_discovery_loop_subject_resolution_optional: true,
+        previous_discovery_entity_candidates: ["\u041d\u0414\u0421"]
+      }
+    });
+
+    expect(result.adapter_status).toBe("ready");
+    expect(result.should_run_discovery).toBe(true);
+    expect(result.semantic_data_need).toBe("movement evidence");
+    expect(result.turn_meaning_ref).toMatchObject({
+      asked_domain_family: "movements",
+      asked_action_family: "list_movements",
+      metadata_scope_hint: "\u041d\u0414\u0421",
+      subject_resolution_optional: true,
+      explicit_organization_scope: orgName,
+      unsupported_but_understood_family: "movement_evidence",
+      stale_replay_forbidden: true
+    });
+    expect(result.turn_meaning_ref?.explicit_entity_candidates).toBeUndefined();
+    expect(result.data_need_graph?.clarification_gaps).toEqual(["period"]);
+  });
 });
@@ -1647,4 +1647,156 @@ describe("assistantTransitionPolicy", () => {
      period_to: "2017-05-31"
    });
  });
+  it("carries metadata-scoped subjectless loop state through follow-up context", () => {
+    const policy = buildPolicy({
+      findLastAddressAssistantItem: () => ({
+        text: "\u043d\u0443\u0436\u043d\u044b \u043e\u0440\u0433\u0430\u043d\u0438\u0437\u0430\u0446\u0438\u044f \u0438 \u043f\u0435\u0440\u0438\u043e\u0434",
+        debug: {
+          execution_lane: "living_chat",
+          mcp_discovery_response_applied: true,
+          assistant_mcp_discovery_entry_point_v1: {
+            schema_version: "assistant_mcp_discovery_runtime_entry_point_v1",
+            entry_status: "bridge_executed",
+            turn_input: {
+              turn_meaning_ref: {
+                asked_domain_family: "movements",
+                asked_action_family: "list_movements",
+                metadata_scope_hint: "\u041d\u0414\u0421",
+                subject_resolution_optional: true,
+                unsupported_but_understood_family: "movement_evidence",
+                stale_replay_forbidden: true
+              }
+            },
+            bridge: {
+              bridge_status: "needs_clarification",
+              business_fact_answer_allowed: false,
+              pilot: {
+                pilot_scope: "counterparty_movement_evidence_query_movements_v1"
+              },
+              loop_state: {
+                schema_version: "assistant_mcp_discovery_loop_state_v1",
+                policy_owner: "assistantMcpDiscoveryRuntimeBridge",
+                loop_status: "awaiting_clarification",
+                selected_chain_id: "movement_evidence",
+                pilot_scope: "counterparty_movement_evidence_query_movements_v1",
+                asked_domain_family: "movements",
+                asked_action_family: "list_movements",
+                unsupported_but_understood_family: "movement_evidence",
+                ranking_need: null,
+                pending_axes: ["organization", "period"],
+                provided_axes: [],
+                explicit_entity_candidates: [],
+                metadata_scope_hint: "\u041d\u0414\u0421",
+                subject_resolution_optional: true,
+                explicit_organization_scope: null,
+                explicit_date_scope: null
+              },
+              answer_draft: {
+                answer_mode: "needs_clarification"
+              }
+            }
+          }
+        }
+      }),
+      hasAddressFollowupContextSignal: () => true,
+      hasReferentialPointer: () => false,
+      resolveAddressIntent: () => ({ intent: "unknown" }),
+      resolveAddressIntentFamily: () => null,
+      resolveAssistantTurnMeaning: () => null
+    });
+
+    const carryover = policy.resolveAddressFollowupCarryoverContext(
+      "\u043f\u043e \u041e\u041e\u041e \u0410\u043b\u044c\u0442\u0435\u0440\u043d\u0430\u0442\u0438\u0432\u0430 \u041f\u043b\u044e\u0441",
+      [{ kind: "assistant", text: "\u043d\u0443\u0436\u043d\u044b \u043e\u0440\u0433\u0430\u043d\u0438\u0437\u0430\u0446\u0438\u044f \u0438 \u043f\u0435\u0440\u0438\u043e\u0434" }],
+      "\u043f\u043e \u041e\u041e\u041e \u0410\u043b\u044c\u0442\u0435\u0440\u043d\u0430\u0442\u0438\u0432\u0430 \u041f\u043b\u044e\u0441",
+      { predecomposeContract: { intent: "unknown" } },
+      null
+    );
+
+    expect(carryover?.followupContext?.previous_discovery_loop_selected_chain_id).toBe("movement_evidence");
+    expect(carryover?.followupContext?.previous_discovery_loop_pending_axes).toEqual([
+      "organization",
+      "period"
+    ]);
+    expect(carryover?.followupContext?.previous_discovery_loop_metadata_scope_hint).toBe(
+      "\u041d\u0414\u0421"
+    );
+    expect(carryover?.followupContext?.previous_discovery_loop_subject_resolution_optional).toBe(true);
+  });
+  it("does not backfill metadata scope into counterparty carryover during lane choice follow-up", () => {
+    const policy = buildPolicy({
+      findLastAddressAssistantItem: () => ({
+        text: "\u0443\u0442\u043e\u0447\u043d\u0438\u0442\u0435: \u043f\u043e \u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442\u0430\u043c \u0438\u043b\u0438 \u043f\u043e \u0434\u0432\u0438\u0436\u0435\u043d\u0438\u044f\u043c?",
+        debug: {
+          execution_lane: "living_chat",
+          mcp_discovery_response_applied: true,
+          assistant_mcp_discovery_entry_point_v1: {
+            schema_version: "assistant_mcp_discovery_runtime_entry_point_v1",
+            entry_status: "bridge_executed",
+            turn_input: {
+              turn_meaning_ref: {
+                asked_domain_family: "metadata",
+                asked_action_family: "resolve_next_lane",
+                explicit_entity_candidates: ["\u041d\u0414\u0421"],
+                metadata_scope_hint: "\u041d\u0414\u0421",
+                metadata_ambiguity_entity_sets: [
+                  "\u0414\u043e\u043a\u0443\u043c\u0435\u043d\u0442",
+                  "\u0420\u0435\u0433\u0438\u0441\u0442\u0440\u041d\u0430\u043a\u043e\u043f\u043b\u0435\u043d\u0438\u044f"
+                ],
+                unsupported_but_understood_family: "metadata_lane_choice_clarification",
+                stale_replay_forbidden: true
+              }
+            },
+            bridge: {
+              bridge_status: "needs_clarification",
+              business_fact_answer_allowed: false,
+              pilot: {
+                pilot_scope: "metadata_inspection_v1"
+              },
+              loop_state: {
+                schema_version: "assistant_mcp_discovery_loop_state_v1",
+                policy_owner: "assistantMcpDiscoveryRuntimeBridge",
+                loop_status: "awaiting_clarification",
+                selected_chain_id: "metadata_lane_clarification",
+                pilot_scope: "metadata_inspection_v1",
+                asked_domain_family: "metadata",
+                asked_action_family: "resolve_next_lane",
+                unsupported_but_understood_family: "metadata_lane_choice_clarification",
+                ranking_need: null,
+                pending_axes: ["lane_family_choice"],
+                provided_axes: [],
+                explicit_entity_candidates: ["\u041d\u0414\u0421"],
+                metadata_scope_hint: "\u041d\u0414\u0421",
+                subject_resolution_optional: false,
+                explicit_organization_scope: null,
+                explicit_date_scope: null
+              },
+              answer_draft: {
+                answer_mode: "needs_clarification"
+              }
+            }
+          }
+        }
+      }),
+      hasAddressFollowupContextSignal: () => true,
+      hasReferentialPointer: () => false,
+      resolveAddressIntent: () => ({ intent: "unknown" }),
+      resolveAddressIntentFamily: () => null,
+      resolveAssistantTurnMeaning: () => null
+    });
+
+    const carryover = policy.resolveAddressFollowupCarryoverContext(
+      "\u043f\u043e \u0434\u0432\u0438\u0436\u0435\u043d\u0438\u044f\u043c",
+      [{ kind: "assistant", text: "\u0443\u0442\u043e\u0447\u043d\u0438\u0442\u0435: \u043f\u043e \u0434\u043e\u043a\u0443\u043c\u0435\u043d\u0442\u0430\u043c \u0438\u043b\u0438 \u043f\u043e \u0434\u0432\u0438\u0436\u0435\u043d\u0438\u044f\u043c?" }],
+      "\u043f\u043e \u0434\u0432\u0438\u0436\u0435\u043d\u0438\u044f\u043c",
+      { predecomposeContract: { intent: "unknown" } },
+      null
+    );
+
+    expect(carryover?.followupContext?.previous_filters?.counterparty).toBeUndefined();
+    expect(carryover?.followupContext?.previous_anchor_type).toBeUndefined();
+    expect(carryover?.followupContext?.previous_anchor_value).toBeNull();
+    expect(carryover?.followupContext?.previous_discovery_entity_candidates).toEqual(["\u041d\u0414\u0421"]);
+    expect(carryover?.followupContext?.previous_discovery_pilot_scope).toBe("metadata_inspection_v1");
+  });
 });