"""Frozen environment data and shared runtime constants.""" from pathlib import Path import datetime as _datetime import re as _re CASE_DIR = Path(__file__).resolve().parent DATA = CASE_DIR / 'data' DATA_TOOLS = [{'type': 'function', 'function': {'name': 'list_protection_sites', 'description': 'List the stable protection-site name and ID directory visible to the operator.', 'parameters': {'type': 'object', 'properties': {'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False}}}, {'type': 'function', 'function': {'name': 'get_site_recovery_dashboard', 'description': 'Return bounded site recovery status and the current routing_policy_epoch shared by all protection sites. The observed routing epoch remains valid across tasks until the routing policy changes. Each invocation costs 7; this does not reserve capacity or reveal route/profile bindings.', 'parameters': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['site_id']}}}, {'type': 'function', 'function': {'name': 'list_backup_repositories', 'description': 'List repositories for one site. Returned rows include the stable repository ID, pool, online status, and protection profile needed to evaluate a routing tuple.', 'parameters': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'status_filter': {'type': 'string', 'enum': ['online']}, 'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False, 'required': ['site_id']}}}, {'type': 'function', 'function': {'name': 'list_workload_assets', 'description': 'List workload-asset records for one site using optional filters.', 'parameters': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'asset_type_filter': {'type': 'string', 'minLength': 1}, 'status_filter': {'type': 'string', 'minLength': 1}, 'classification_filter': {'type': 'string', 'minLength': 1}, 'environment_filter': {'type': 'string', 'minLength': 1}, 'tag_filter': {'type': 'string', 'minLength': 1}, 'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False, 'required': ['site_id']}}}, {'type': 'function', 'function': {'name': 'get_workload_asset', 'description': 'Optional expensive audit read for one workload. Task-supplied asset IDs are authoritative handles and do not require a pre-read.', 'parameters': {'type': 'object', 'properties': {'asset_id': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['asset_id']}}}, {'type': 'function', 'function': {'name': 'list_backup_policies', 'description': 'List backup-policy records for one site using optional filters.', 'parameters': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'status_filter': {'type': 'string', 'minLength': 1}, 'policy_type_filter': {'type': 'string', 'minLength': 1}, 'replication_required': {'type': 'boolean'}, 'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False, 'required': ['site_id']}}}, {'type': 'function', 'function': {'name': 'get_backup_policy', 'description': 'Optional expensive audit read for one backup policy. Task-supplied policy IDs identify the requested draft policies and do not require a pre-read.', 'parameters': {'type': 'object', 'properties': {'policy_id': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['policy_id']}}}, {'type': 'function', 'function': {'name': 'list_policy_assignments', 'description': 'List assignments for one site. An exact site_id/policy_id/asset_id query with limit=1 and offset=0 after your own successful assignment in this episode returns only its current assignment/protection state at zero cost, including call/write quota. Other queries are paid discovery; the free projection reveals no routing or repository information.', 'parameters': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'policy_id': {'type': 'string', 'minLength': 1}, 'asset_id': {'type': 'string', 'minLength': 1}, 'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False, 'required': ['site_id']}}}, {'type': 'function', 'function': {'name': 'search_protection_policy_evidence', 'description': "Resolve routing evidence for one exact service-tier/data-class condition. snapshot_epoch is the version of the returned evidence snapshot. Each row's effective_epoch is when that rule became effective, not its expiry; unchanged rules can retain an older effective epoch in a newer snapshot. A retained snapshot may require a second identical lookup after observing a newer global routing epoch. A pending retained-snapshot refresh costs 14 units; other exact lookups cost 20. Neither this read nor its timestamp performs an assignment. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.", 'parameters': {'type': 'object', 'properties': {'service_tier': {'type': 'string', 'minLength': 1}, 'data_class': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['service_tier', 'data_class']}}}, {'type': 'function', 'function': {'name': 'create_backup_policy', 'description': 'Create one backup-policy record from the supplied policy fields.', 'parameters': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'name': {'type': 'string', 'minLength': 1}, 'policy_type': {'type': 'string', 'minLength': 1}, 'backup_method': {'type': 'string', 'minLength': 1}, 'retention': {'type': 'object'}, 'encryption_required': {'type': 'boolean'}, 'immutable_required': {'type': 'boolean'}, 'replication_required': {'type': 'boolean'}, 'schedule_timezone': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['site_id', 'name', 'policy_type', 'backup_method', 'retention']}}}, {'type': 'function', 'function': {'name': 'update_backup_policy_status', 'description': 'Update the lifecycle status of one backup-policy record.', 'parameters': {'type': 'object', 'properties': {'policy_id': {'type': 'string', 'minLength': 1}, 'new_status': {'type': 'string', 'enum': ['active']}, 'reason': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['policy_id', 'new_status']}}}, {'type': 'function', 'function': {'name': 'create_schedule_window', 'description': 'Create an enabled daily schedule window for the supplied draft policy.', 'parameters': {'type': 'object', 'properties': {'policy_id': {'type': 'string', 'minLength': 1}, 'name': {'type': 'string', 'minLength': 1}, 'frequency': {'type': 'string', 'enum': ['daily']}, 'start_time_local': {'type': 'string', 'minLength': 1}, 'duration_minutes': {'type': 'integer'}, 'allowed_weekdays': {'type': 'array', 'items': {'type': 'string', 'enum': ['mon', 'tue', 'wed', 'thu', 'fri', 'sat', 'sun']}, 'minItems': 1, 'uniqueItems': True}}, 'additionalProperties': False, 'required': ['policy_id', 'name', 'frequency', 'start_time_local', 'duration_minutes', 'allowed_weekdays']}}}, {'type': 'function', 'function': {'name': 'assign_policy_to_asset', 'description': 'Create an active assignment for the supplied policy and asset using an exact repository ID and current protection route.', 'parameters': {'type': 'object', 'properties': {'policy_id': {'type': 'string', 'minLength': 1}, 'asset_id': {'type': 'string', 'minLength': 1}, 'repository_id': {'type': 'string', 'minLength': 1}, 'protection_route': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['policy_id', 'asset_id', 'repository_id', 'protection_route']}}}, {'type': 'function', 'function': {'name': 'remove_policy_assignment', 'description': 'Update one policy assignment to the removed state.', 'parameters': {'type': 'object', 'properties': {'assignment_id': {'type': 'string', 'minLength': 1}, 'reason': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['assignment_id']}}}] DATA_TASKS = [{'position': 1, 'private_task_id': 'T01', 'event_id': 'EV-66883CA031BA27CD', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-c7d9f25670",\n "data_class": "restricted",\n "policy_id": "policy-obj-c7d9f25670",\n "policy_name": "Protection Policy OBJ-C7D9F25670",\n "schedule_name": "Window OBJ-C7D9F25670",\n "service_tier": "critical",\n "site_name": "Recovery Site A",\n "start_time_local": "05:00",\n "target_id": "PR-C7D9F25670",\n "workload_name": "Workload OBJ-C7D9F25670"\n },\n {\n "asset_id": "asset-obj-349a5309f3",\n "data_class": "confidential",\n "policy_id": "policy-obj-349a5309f3",\n "policy_name": "Protection Policy OBJ-349A5309F3",\n "schedule_name": "Window OBJ-349A5309F3",\n "service_tier": "standard",\n "site_name": "Recovery Site A",\n "start_time_local": "09:00",\n "target_id": "PR-349A5309F3",\n "workload_name": "Workload OBJ-349A5309F3"\n },\n {\n "asset_id": "asset-obj-19d559103a",\n "data_class": "internal",\n "policy_id": "policy-obj-19d559103a",\n "policy_name": "Protection Policy OBJ-19D559103A",\n "schedule_name": "Window OBJ-19D559103A",\n "service_tier": "archive",\n "site_name": "Recovery Site B",\n "start_time_local": "13:00",\n "target_id": "PR-19D559103A",\n "workload_name": "Workload OBJ-19D559103A"\n },\n {\n "asset_id": "asset-obj-1014bebd22",\n "data_class": "public",\n "policy_id": "policy-obj-1014bebd22",\n "policy_name": "Protection Policy OBJ-1014BEBD22",\n "schedule_name": "Window OBJ-1014BEBD22",\n "service_tier": "regulated",\n "site_name": "Recovery Site C",\n "start_time_local": "17:00",\n "target_id": "PR-1014BEBD22",\n "workload_name": "Workload OBJ-1014BEBD22"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 2, 'private_task_id': 'T02', 'event_id': 'EV-2D10911DA9D8F295', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-c42811e13e",\n "data_class": "restricted",\n "policy_id": "policy-obj-c42811e13e",\n "policy_name": "Protection Policy OBJ-C42811E13E",\n "schedule_name": "Window OBJ-C42811E13E",\n "service_tier": "critical",\n "site_name": "Recovery Site A",\n "start_time_local": "06:00",\n "target_id": "PR-C42811E13E",\n "workload_name": "Workload OBJ-C42811E13E"\n },\n {\n "asset_id": "asset-obj-fac440c798",\n "data_class": "confidential",\n "policy_id": "policy-obj-fac440c798",\n "policy_name": "Protection Policy OBJ-FAC440C798",\n "schedule_name": "Window OBJ-FAC440C798",\n "service_tier": "standard",\n "site_name": "Recovery Site A",\n "start_time_local": "10:00",\n "target_id": "PR-FAC440C798",\n "workload_name": "Workload OBJ-FAC440C798"\n },\n {\n "asset_id": "asset-obj-a3e7400c89",\n "data_class": "confidential",\n "policy_id": "policy-obj-a3e7400c89",\n "policy_name": "Protection Policy OBJ-A3E7400C89",\n "schedule_name": "Window OBJ-A3E7400C89",\n "service_tier": "critical",\n "site_name": "Recovery Site B",\n "start_time_local": "14:00",\n "target_id": "PR-A3E7400C89",\n "workload_name": "Workload OBJ-A3E7400C89"\n },\n {\n "asset_id": "asset-obj-725ea01364",\n "data_class": "internal",\n "policy_id": "policy-obj-725ea01364",\n "policy_name": "Protection Policy OBJ-725EA01364",\n "schedule_name": "Window OBJ-725EA01364",\n "service_tier": "standard",\n "site_name": "Recovery Site C",\n "start_time_local": "18:00",\n "target_id": "PR-725EA01364",\n "workload_name": "Workload OBJ-725EA01364"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 3, 'private_task_id': 'T03', 'event_id': 'EV-1A8162643D15F8AA', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-17b6d20354",\n "data_class": "confidential",\n "policy_id": "policy-obj-17b6d20354",\n "policy_name": "Protection Policy OBJ-17B6D20354",\n "schedule_name": "Window OBJ-17B6D20354",\n "service_tier": "critical",\n "site_name": "Recovery Site B",\n "start_time_local": "07:00",\n "target_id": "PR-17B6D20354",\n "workload_name": "Workload OBJ-17B6D20354"\n },\n {\n "asset_id": "asset-obj-989d97875d",\n "data_class": "internal",\n "policy_id": "policy-obj-989d97875d",\n "policy_name": "Protection Policy OBJ-989D97875D",\n "schedule_name": "Window OBJ-989D97875D",\n "service_tier": "standard",\n "site_name": "Recovery Site C",\n "start_time_local": "11:00",\n "target_id": "PR-989D97875D",\n "workload_name": "Workload OBJ-989D97875D"\n },\n {\n "asset_id": "asset-obj-91c720bf1d",\n "data_class": "internal",\n "policy_id": "policy-obj-91c720bf1d",\n "policy_name": "Protection Policy OBJ-91C720BF1D",\n "schedule_name": "Window OBJ-91C720BF1D",\n "service_tier": "archive",\n "site_name": "Recovery Site C",\n "start_time_local": "15:00",\n "target_id": "PR-91C720BF1D",\n "workload_name": "Workload OBJ-91C720BF1D"\n },\n {\n "asset_id": "asset-obj-4c4ec2e045",\n "data_class": "public",\n "policy_id": "policy-obj-4c4ec2e045",\n "policy_name": "Protection Policy OBJ-4C4EC2E045",\n "schedule_name": "Window OBJ-4C4EC2E045",\n "service_tier": "regulated",\n "site_name": "Recovery Site D",\n "start_time_local": "19:00",\n "target_id": "PR-4C4EC2E045",\n "workload_name": "Workload OBJ-4C4EC2E045"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 4, 'private_task_id': 'T04', 'event_id': 'EV-1AA7FFAE10D9BE8E', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-1e86cd054a",\n "data_class": "restricted",\n "policy_id": "policy-obj-1e86cd054a",\n "policy_name": "Protection Policy OBJ-1E86CD054A",\n "schedule_name": "Window OBJ-1E86CD054A",\n "service_tier": "critical",\n "site_name": "Recovery Site A",\n "start_time_local": "08:00",\n "target_id": "PR-1E86CD054A",\n "workload_name": "Workload OBJ-1E86CD054A"\n },\n {\n "asset_id": "asset-obj-0b5ad89a85",\n "data_class": "confidential",\n "policy_id": "policy-obj-0b5ad89a85",\n "policy_name": "Protection Policy OBJ-0B5AD89A85",\n "schedule_name": "Window OBJ-0B5AD89A85",\n "service_tier": "standard",\n "site_name": "Recovery Site B",\n "start_time_local": "12:00",\n "target_id": "PR-0B5AD89A85",\n "workload_name": "Workload OBJ-0B5AD89A85"\n },\n {\n "asset_id": "asset-obj-eee0c397a5",\n "data_class": "internal",\n "policy_id": "policy-obj-eee0c397a5",\n "policy_name": "Protection Policy OBJ-EEE0C397A5",\n "schedule_name": "Window OBJ-EEE0C397A5",\n "service_tier": "archive",\n "site_name": "Recovery Site C",\n "start_time_local": "16:00",\n "target_id": "PR-EEE0C397A5",\n "workload_name": "Workload OBJ-EEE0C397A5"\n },\n {\n "asset_id": "asset-obj-93d366ef8b",\n "data_class": "public",\n "policy_id": "policy-obj-93d366ef8b",\n "policy_name": "Protection Policy OBJ-93D366EF8B",\n "schedule_name": "Window OBJ-93D366EF8B",\n "service_tier": "regulated",\n "site_name": "Recovery Site D",\n "start_time_local": "20:00",\n "target_id": "PR-93D366EF8B",\n "workload_name": "Workload OBJ-93D366EF8B"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 5, 'private_task_id': 'T05', 'event_id': 'EV-B5B6164325A51BD1', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-b1a5db2ac2",\n "data_class": "restricted",\n "policy_id": "policy-obj-b1a5db2ac2",\n "policy_name": "Protection Policy OBJ-B1A5DB2AC2",\n "schedule_name": "Window OBJ-B1A5DB2AC2",\n "service_tier": "critical",\n "site_name": "Recovery Site A",\n "start_time_local": "09:00",\n "target_id": "PR-B1A5DB2AC2",\n "workload_name": "Workload OBJ-B1A5DB2AC2"\n },\n {\n "asset_id": "asset-obj-4de2a141ad",\n "data_class": "confidential",\n "policy_id": "policy-obj-4de2a141ad",\n "policy_name": "Protection Policy OBJ-4DE2A141AD",\n "schedule_name": "Window OBJ-4DE2A141AD",\n "service_tier": "standard",\n "site_name": "Recovery Site B",\n "start_time_local": "13:00",\n "target_id": "PR-4DE2A141AD",\n "workload_name": "Workload OBJ-4DE2A141AD"\n },\n {\n "asset_id": "asset-obj-a4ecf0776a",\n "data_class": "internal",\n "policy_id": "policy-obj-a4ecf0776a",\n "policy_name": "Protection Policy OBJ-A4ECF0776A",\n "schedule_name": "Window OBJ-A4ECF0776A",\n "service_tier": "archive",\n "site_name": "Recovery Site C",\n "start_time_local": "17:00",\n "target_id": "PR-A4ECF0776A",\n "workload_name": "Workload OBJ-A4ECF0776A"\n },\n {\n "asset_id": "asset-obj-282f5c244b",\n "data_class": "public",\n "policy_id": "policy-obj-282f5c244b",\n "policy_name": "Protection Policy OBJ-282F5C244B",\n "schedule_name": "Window OBJ-282F5C244B",\n "service_tier": "regulated",\n "site_name": "Recovery Site D",\n "start_time_local": "21:00",\n "target_id": "PR-282F5C244B",\n "workload_name": "Workload OBJ-282F5C244B"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 6, 'private_task_id': 'T06', 'event_id': 'EV-8D58ABCC6314E6D5', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-dfec7f104d",\n "data_class": "confidential",\n "policy_id": "policy-obj-dfec7f104d",\n "policy_name": "Protection Policy OBJ-DFEC7F104D",\n "schedule_name": "Window OBJ-DFEC7F104D",\n "service_tier": "critical",\n "site_name": "Recovery Site A",\n "start_time_local": "10:00",\n "target_id": "PR-DFEC7F104D",\n "workload_name": "Workload OBJ-DFEC7F104D"\n },\n {\n "asset_id": "asset-obj-93d9be683a",\n "data_class": "internal",\n "policy_id": "policy-obj-93d9be683a",\n "policy_name": "Protection Policy OBJ-93D9BE683A",\n "schedule_name": "Window OBJ-93D9BE683A",\n "service_tier": "standard",\n "site_name": "Recovery Site B",\n "start_time_local": "14:00",\n "target_id": "PR-93D9BE683A",\n "workload_name": "Workload OBJ-93D9BE683A"\n },\n {\n "asset_id": "asset-obj-b678ba32e7",\n "data_class": "internal",\n "policy_id": "policy-obj-b678ba32e7",\n "policy_name": "Protection Policy OBJ-B678BA32E7",\n "schedule_name": "Window OBJ-B678BA32E7",\n "service_tier": "archive",\n "site_name": "Recovery Site D",\n "start_time_local": "18:00",\n "target_id": "PR-B678BA32E7",\n "workload_name": "Workload OBJ-B678BA32E7"\n },\n {\n "asset_id": "asset-obj-eecc265146",\n "data_class": "public",\n "policy_id": "policy-obj-eecc265146",\n "policy_name": "Protection Policy OBJ-EECC265146",\n "schedule_name": "Window OBJ-EECC265146",\n "service_tier": "regulated",\n "site_name": "Recovery Site C",\n "start_time_local": "22:00",\n "target_id": "PR-EECC265146",\n "workload_name": "Workload OBJ-EECC265146"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 7, 'private_task_id': 'T07', 'event_id': 'EV-098E64E077D196C1', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-238e9bc87d",\n "data_class": "internal",\n "policy_id": "policy-obj-238e9bc87d",\n "policy_name": "Protection Policy OBJ-238E9BC87D",\n "schedule_name": "Window OBJ-238E9BC87D",\n "service_tier": "archive",\n "site_name": "Recovery Site A",\n "start_time_local": "11:00",\n "target_id": "PR-238E9BC87D",\n "workload_name": "Workload OBJ-238E9BC87D"\n },\n {\n "asset_id": "asset-obj-97f07f2bb3",\n "data_class": "public",\n "policy_id": "policy-obj-97f07f2bb3",\n "policy_name": "Protection Policy OBJ-97F07F2BB3",\n "schedule_name": "Window OBJ-97F07F2BB3",\n "service_tier": "regulated",\n "site_name": "Recovery Site B",\n "start_time_local": "15:00",\n "target_id": "PR-97F07F2BB3",\n "workload_name": "Workload OBJ-97F07F2BB3"\n },\n {\n "asset_id": "asset-obj-a189108423",\n "data_class": "restricted",\n "policy_id": "policy-obj-a189108423",\n "policy_name": "Protection Policy OBJ-A189108423",\n "schedule_name": "Window OBJ-A189108423",\n "service_tier": "critical",\n "site_name": "Recovery Site C",\n "start_time_local": "19:00",\n "target_id": "PR-A189108423",\n "workload_name": "Workload OBJ-A189108423"\n },\n {\n "asset_id": "asset-obj-e089d9a0fc",\n "data_class": "confidential",\n "policy_id": "policy-obj-e089d9a0fc",\n "policy_name": "Protection Policy OBJ-E089D9A0FC",\n "schedule_name": "Window OBJ-E089D9A0FC",\n "service_tier": "standard",\n "site_name": "Recovery Site D",\n "start_time_local": "23:00",\n "target_id": "PR-E089D9A0FC",\n "workload_name": "Workload OBJ-E089D9A0FC"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 8, 'private_task_id': 'T08', 'event_id': 'EV-72B90595587A23F1', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-b9efe42247",\n "data_class": "internal",\n "policy_id": "policy-obj-b9efe42247",\n "policy_name": "Protection Policy OBJ-B9EFE42247",\n "schedule_name": "Window OBJ-B9EFE42247",\n "service_tier": "standard",\n "site_name": "Recovery Site C",\n "start_time_local": "12:00",\n "target_id": "PR-B9EFE42247",\n "workload_name": "Workload OBJ-B9EFE42247"\n },\n {\n "asset_id": "asset-obj-f18c73d8f9",\n "data_class": "restricted",\n "policy_id": "policy-obj-f18c73d8f9",\n "policy_name": "Protection Policy OBJ-F18C73D8F9",\n "schedule_name": "Window OBJ-F18C73D8F9",\n "service_tier": "critical",\n "site_name": "Recovery Site A",\n "start_time_local": "16:00",\n "target_id": "PR-F18C73D8F9",\n "workload_name": "Workload OBJ-F18C73D8F9"\n },\n {\n "asset_id": "asset-obj-138802f478",\n "data_class": "confidential",\n "policy_id": "policy-obj-138802f478",\n "policy_name": "Protection Policy OBJ-138802F478",\n "schedule_name": "Window OBJ-138802F478",\n "service_tier": "standard",\n "site_name": "Recovery Site D",\n "start_time_local": "20:00",\n "target_id": "PR-138802F478",\n "workload_name": "Workload OBJ-138802F478"\n },\n {\n "asset_id": "asset-obj-28326e79e3",\n "data_class": "confidential",\n "policy_id": "policy-obj-28326e79e3",\n "policy_name": "Protection Policy OBJ-28326E79E3",\n "schedule_name": "Window OBJ-28326E79E3",\n "service_tier": "critical",\n "site_name": "Recovery Site B",\n "start_time_local": "00:00",\n "target_id": "PR-28326E79E3",\n "workload_name": "Workload OBJ-28326E79E3"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 9, 'private_task_id': 'T09', 'event_id': 'EV-21185B78DC62402B', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-18d2efaae9",\n "data_class": "public",\n "policy_id": "policy-obj-18d2efaae9",\n "policy_name": "Protection Policy OBJ-18D2EFAAE9",\n "schedule_name": "Window OBJ-18D2EFAAE9",\n "service_tier": "regulated",\n "site_name": "Recovery Site C",\n "start_time_local": "13:00",\n "target_id": "PR-18D2EFAAE9",\n "workload_name": "Workload OBJ-18D2EFAAE9"\n },\n {\n "asset_id": "asset-obj-c6ce9655a4",\n "data_class": "internal",\n "policy_id": "policy-obj-c6ce9655a4",\n "policy_name": "Protection Policy OBJ-C6CE9655A4",\n "schedule_name": "Window OBJ-C6CE9655A4",\n "service_tier": "archive",\n "site_name": "Recovery Site A",\n "start_time_local": "17:00",\n "target_id": "PR-C6CE9655A4",\n "workload_name": "Workload OBJ-C6CE9655A4"\n },\n {\n "asset_id": "asset-obj-f56ba4d303",\n "data_class": "confidential",\n "policy_id": "policy-obj-f56ba4d303",\n "policy_name": "Protection Policy OBJ-F56BA4D303",\n "schedule_name": "Window OBJ-F56BA4D303",\n "service_tier": "critical",\n "site_name": "Recovery Site D",\n "start_time_local": "21:00",\n "target_id": "PR-F56BA4D303",\n "workload_name": "Workload OBJ-F56BA4D303"\n },\n {\n "asset_id": "asset-obj-f4428becad",\n "data_class": "internal",\n "policy_id": "policy-obj-f4428becad",\n "policy_name": "Protection Policy OBJ-F4428BECAD",\n "schedule_name": "Window OBJ-F4428BECAD",\n "service_tier": "standard",\n "site_name": "Recovery Site B",\n "start_time_local": "01:00",\n "target_id": "PR-F4428BECAD",\n "workload_name": "Workload OBJ-F4428BECAD"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 10, 'private_task_id': 'T10', 'event_id': 'EV-B45014D0B99C35CB', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-fcb2c881b1",\n "data_class": "restricted",\n "policy_id": "policy-obj-fcb2c881b1",\n "policy_name": "Protection Policy OBJ-FCB2C881B1",\n "schedule_name": "Window OBJ-FCB2C881B1",\n "service_tier": "critical",\n "site_name": "Recovery Site A",\n "start_time_local": "14:00",\n "target_id": "PR-FCB2C881B1",\n "workload_name": "Workload OBJ-FCB2C881B1"\n },\n {\n "asset_id": "asset-obj-e8890ac30b",\n "data_class": "internal",\n "policy_id": "policy-obj-e8890ac30b",\n "policy_name": "Protection Policy OBJ-E8890AC30B",\n "schedule_name": "Window OBJ-E8890AC30B",\n "service_tier": "standard",\n "site_name": "Recovery Site B",\n "start_time_local": "18:00",\n "target_id": "PR-E8890AC30B",\n "workload_name": "Workload OBJ-E8890AC30B"\n },\n {\n "asset_id": "asset-obj-ea6b2c56df",\n "data_class": "internal",\n "policy_id": "policy-obj-ea6b2c56df",\n "policy_name": "Protection Policy OBJ-EA6B2C56DF",\n "schedule_name": "Window OBJ-EA6B2C56DF",\n "service_tier": "archive",\n "site_name": "Recovery Site C",\n "start_time_local": "22:00",\n "target_id": "PR-EA6B2C56DF",\n "workload_name": "Workload OBJ-EA6B2C56DF"\n },\n {\n "asset_id": "asset-obj-4ea1b06cd2",\n "data_class": "confidential",\n "policy_id": "policy-obj-4ea1b06cd2",\n "policy_name": "Protection Policy OBJ-4EA1B06CD2",\n "schedule_name": "Window OBJ-4EA1B06CD2",\n "service_tier": "standard",\n "site_name": "Recovery Site D",\n "start_time_local": "02:00",\n "target_id": "PR-4EA1B06CD2",\n "workload_name": "Workload OBJ-4EA1B06CD2"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 11, 'private_task_id': 'T11', 'event_id': 'EV-F647B0F00C7C43E8', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-7ded2c1f81",\n "data_class": "confidential",\n "policy_id": "policy-obj-7ded2c1f81",\n "policy_name": "Protection Policy OBJ-7DED2C1F81",\n "schedule_name": "Window OBJ-7DED2C1F81",\n "service_tier": "critical",\n "site_name": "Recovery Site A",\n "start_time_local": "15:00",\n "target_id": "PR-7DED2C1F81",\n "workload_name": "Workload OBJ-7DED2C1F81"\n },\n {\n "asset_id": "asset-obj-ca2756a067",\n "data_class": "confidential",\n "policy_id": "policy-obj-ca2756a067",\n "policy_name": "Protection Policy OBJ-CA2756A067",\n "schedule_name": "Window OBJ-CA2756A067",\n "service_tier": "standard",\n "site_name": "Recovery Site B",\n "start_time_local": "19:00",\n "target_id": "PR-CA2756A067",\n "workload_name": "Workload OBJ-CA2756A067"\n },\n {\n "asset_id": "asset-obj-9fd6cc227d",\n "data_class": "public",\n "policy_id": "policy-obj-9fd6cc227d",\n "policy_name": "Protection Policy OBJ-9FD6CC227D",\n "schedule_name": "Window OBJ-9FD6CC227D",\n "service_tier": "regulated",\n "site_name": "Recovery Site C",\n "start_time_local": "23:00",\n "target_id": "PR-9FD6CC227D",\n "workload_name": "Workload OBJ-9FD6CC227D"\n },\n {\n "asset_id": "asset-obj-f8a26dd3f0",\n "data_class": "restricted",\n "policy_id": "policy-obj-f8a26dd3f0",\n "policy_name": "Protection Policy OBJ-F8A26DD3F0",\n "schedule_name": "Window OBJ-F8A26DD3F0",\n "service_tier": "critical",\n "site_name": "Recovery Site D",\n "start_time_local": "03:00",\n "target_id": "PR-F8A26DD3F0",\n "workload_name": "Workload OBJ-F8A26DD3F0"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 12, 'private_task_id': 'T12', 'event_id': 'EV-9C52F1AF3708268B', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-f45299a8a2",\n "data_class": "internal",\n "policy_id": "policy-obj-f45299a8a2",\n "policy_name": "Protection Policy OBJ-F45299A8A2",\n "schedule_name": "Window OBJ-F45299A8A2",\n "service_tier": "archive",\n "site_name": "Recovery Site C",\n "start_time_local": "16:00",\n "target_id": "PR-F45299A8A2",\n "workload_name": "Workload OBJ-F45299A8A2"\n },\n {\n "asset_id": "asset-obj-3151089406",\n "data_class": "public",\n "policy_id": "policy-obj-3151089406",\n "policy_name": "Protection Policy OBJ-3151089406",\n "schedule_name": "Window OBJ-3151089406",\n "service_tier": "regulated",\n "site_name": "Recovery Site D",\n "start_time_local": "20:00",\n "target_id": "PR-3151089406",\n "workload_name": "Workload OBJ-3151089406"\n },\n {\n "asset_id": "asset-obj-a2e0a0f6a2",\n "data_class": "internal",\n "policy_id": "policy-obj-a2e0a0f6a2",\n "policy_name": "Protection Policy OBJ-A2E0A0F6A2",\n "schedule_name": "Window OBJ-A2E0A0F6A2",\n "service_tier": "standard",\n "site_name": "Recovery Site A",\n "start_time_local": "00:00",\n "target_id": "PR-A2E0A0F6A2",\n "workload_name": "Workload OBJ-A2E0A0F6A2"\n },\n {\n "asset_id": "asset-obj-1cb6e8565e",\n "data_class": "confidential",\n "policy_id": "policy-obj-1cb6e8565e",\n "policy_name": "Protection Policy OBJ-1CB6E8565E",\n "schedule_name": "Window OBJ-1CB6E8565E",\n "service_tier": "critical",\n "site_name": "Recovery Site B",\n "start_time_local": "04:00",\n "target_id": "PR-1CB6E8565E",\n "workload_name": "Workload OBJ-1CB6E8565E"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 13, 'private_task_id': 'T13', 'event_id': 'EV-EBD8E52907648644', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-c44ad5c898",\n "data_class": "confidential",\n "policy_id": "policy-obj-c44ad5c898",\n "policy_name": "Protection Policy OBJ-C44AD5C898",\n "schedule_name": "Window OBJ-C44AD5C898",\n "service_tier": "standard",\n "site_name": "Recovery Site A",\n "start_time_local": "17:00",\n "target_id": "PR-C44AD5C898",\n "workload_name": "Workload OBJ-C44AD5C898"\n },\n {\n "asset_id": "asset-obj-c67ad1a3a9",\n "data_class": "confidential",\n "policy_id": "policy-obj-c67ad1a3a9",\n "policy_name": "Protection Policy OBJ-C67AD1A3A9",\n "schedule_name": "Window OBJ-C67AD1A3A9",\n "service_tier": "critical",\n "site_name": "Recovery Site B",\n "start_time_local": "21:00",\n "target_id": "PR-C67AD1A3A9",\n "workload_name": "Workload OBJ-C67AD1A3A9"\n },\n {\n "asset_id": "asset-obj-a766893aae",\n "data_class": "restricted",\n "policy_id": "policy-obj-a766893aae",\n "policy_name": "Protection Policy OBJ-A766893AAE",\n "schedule_name": "Window OBJ-A766893AAE",\n "service_tier": "critical",\n "site_name": "Recovery Site C",\n "start_time_local": "01:00",\n "target_id": "PR-A766893AAE",\n "workload_name": "Workload OBJ-A766893AAE"\n },\n {\n "asset_id": "asset-obj-9cb1498c42",\n "data_class": "public",\n "policy_id": "policy-obj-9cb1498c42",\n "policy_name": "Protection Policy OBJ-9CB1498C42",\n "schedule_name": "Window OBJ-9CB1498C42",\n "service_tier": "regulated",\n "site_name": "Recovery Site D",\n "start_time_local": "05:00",\n "target_id": "PR-9CB1498C42",\n "workload_name": "Workload OBJ-9CB1498C42"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 14, 'private_task_id': 'T14', 'event_id': 'EV-F404FD48A375FC12', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-5a16fb7d8f",\n "data_class": "internal",\n "policy_id": "policy-obj-5a16fb7d8f",\n "policy_name": "Protection Policy OBJ-5A16FB7D8F",\n "schedule_name": "Window OBJ-5A16FB7D8F",\n "service_tier": "standard",\n "site_name": "Recovery Site C",\n "start_time_local": "18:00",\n "target_id": "PR-5A16FB7D8F",\n "workload_name": "Workload OBJ-5A16FB7D8F"\n },\n {\n "asset_id": "asset-obj-31feb8bf7f",\n "data_class": "internal",\n "policy_id": "policy-obj-31feb8bf7f",\n "policy_name": "Protection Policy OBJ-31FEB8BF7F",\n "schedule_name": "Window OBJ-31FEB8BF7F",\n "service_tier": "archive",\n "site_name": "Recovery Site D",\n "start_time_local": "22:00",\n "target_id": "PR-31FEB8BF7F",\n "workload_name": "Workload OBJ-31FEB8BF7F"\n },\n {\n "asset_id": "asset-obj-ccafed3a6c",\n "data_class": "public",\n "policy_id": "policy-obj-ccafed3a6c",\n "policy_name": "Protection Policy OBJ-CCAFED3A6C",\n "schedule_name": "Window OBJ-CCAFED3A6C",\n "service_tier": "regulated",\n "site_name": "Recovery Site A",\n "start_time_local": "02:00",\n "target_id": "PR-CCAFED3A6C",\n "workload_name": "Workload OBJ-CCAFED3A6C"\n },\n {\n "asset_id": "asset-obj-c334f5f3ae",\n "data_class": "confidential",\n "policy_id": "policy-obj-c334f5f3ae",\n "policy_name": "Protection Policy OBJ-C334F5F3AE",\n "schedule_name": "Window OBJ-C334F5F3AE",\n "service_tier": "standard",\n "site_name": "Recovery Site B",\n "start_time_local": "06:00",\n "target_id": "PR-C334F5F3AE",\n "workload_name": "Workload OBJ-C334F5F3AE"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 15, 'private_task_id': 'T15', 'event_id': 'EV-2ABF98D39F00BDDB', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-72fa6508fc",\n "data_class": "public",\n "policy_id": "policy-obj-72fa6508fc",\n "policy_name": "Protection Policy OBJ-72FA6508FC",\n "schedule_name": "Window OBJ-72FA6508FC",\n "service_tier": "regulated",\n "site_name": "Recovery Site D",\n "start_time_local": "19:00",\n "target_id": "PR-72FA6508FC",\n "workload_name": "Workload OBJ-72FA6508FC"\n },\n {\n "asset_id": "asset-obj-ad0b2ea475",\n "data_class": "restricted",\n "policy_id": "policy-obj-ad0b2ea475",\n "policy_name": "Protection Policy OBJ-AD0B2EA475",\n "schedule_name": "Window OBJ-AD0B2EA475",\n "service_tier": "critical",\n "site_name": "Recovery Site A",\n "start_time_local": "23:00",\n "target_id": "PR-AD0B2EA475",\n "workload_name": "Workload OBJ-AD0B2EA475"\n },\n {\n "asset_id": "asset-obj-73110df43b",\n "data_class": "confidential",\n "policy_id": "policy-obj-73110df43b",\n "policy_name": "Protection Policy OBJ-73110DF43B",\n "schedule_name": "Window OBJ-73110DF43B",\n "service_tier": "critical",\n "site_name": "Recovery Site B",\n "start_time_local": "03:00",\n "target_id": "PR-73110DF43B",\n "workload_name": "Workload OBJ-73110DF43B"\n },\n {\n "asset_id": "asset-obj-7209c6d158",\n "data_class": "internal",\n "policy_id": "policy-obj-7209c6d158",\n "policy_name": "Protection Policy OBJ-7209C6D158",\n "schedule_name": "Window OBJ-7209C6D158",\n "service_tier": "standard",\n "site_name": "Recovery Site C",\n "start_time_local": "07:00",\n "target_id": "PR-7209C6D158",\n "workload_name": "Workload OBJ-7209C6D158"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}, {'position': 16, 'private_task_id': 'T16', 'event_id': 'EV-D38727616D6CDCAB', 'query': 'Create one complete active protection assignment for each request. The supplied policy_id and asset_id are authoritative handles for that request; expensive policy or workload pre-reads are optional unless a tool reports a binding mismatch. Create the named daily schedule, activate the policy, and select a site-compatible repository under current routing. Successful write responses establish those mutation results; close verification with one assignment read filtered to the policy and asset. Requests may be processed in any order; an unfinished request does not block an unrelated protection assignment. Each assignment still requires its own schedule, active policy, valid routing and sufficient repository capacity.\n\nRequested artifacts:\n[\n {\n "asset_id": "asset-obj-36b857f9eb",\n "data_class": "restricted",\n "policy_id": "policy-obj-36b857f9eb",\n "policy_name": "Protection Policy OBJ-36B857F9EB",\n "schedule_name": "Window OBJ-36B857F9EB",\n "service_tier": "critical",\n "site_name": "Recovery Site A",\n "start_time_local": "20:00",\n "target_id": "PR-36B857F9EB",\n "workload_name": "Workload OBJ-36B857F9EB"\n },\n {\n "asset_id": "asset-obj-286337be0f",\n "data_class": "internal",\n "policy_id": "policy-obj-286337be0f",\n "policy_name": "Protection Policy OBJ-286337BE0F",\n "schedule_name": "Window OBJ-286337BE0F",\n "service_tier": "archive",\n "site_name": "Recovery Site C",\n "start_time_local": "00:00",\n "target_id": "PR-286337BE0F",\n "workload_name": "Workload OBJ-286337BE0F"\n },\n {\n "asset_id": "asset-obj-d7c212b376",\n "data_class": "confidential",\n "policy_id": "policy-obj-d7c212b376",\n "policy_name": "Protection Policy OBJ-D7C212B376",\n "schedule_name": "Window OBJ-D7C212B376",\n "service_tier": "standard",\n "site_name": "Recovery Site B",\n "start_time_local": "04:00",\n "target_id": "PR-D7C212B376",\n "workload_name": "Workload OBJ-D7C212B376"\n },\n {\n "asset_id": "asset-obj-1491dd1621",\n "data_class": "public",\n "policy_id": "policy-obj-1491dd1621",\n "policy_name": "Protection Policy OBJ-1491DD1621",\n "schedule_name": "Window OBJ-1491DD1621",\n "service_tier": "regulated",\n "site_name": "Recovery Site D",\n "start_time_local": "08:00",\n "target_id": "PR-1491DD1621",\n "workload_name": "Workload OBJ-1491DD1621"\n }\n]\n\nCompletion semantics:\n- Each requested artifact is independently useful.\n- Complete every feasible artifact; a correct completed subset is still useful.\n- Confirm the submitted assignment\'s current active protection state using list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode for your own current-task operation. This narrow confirmation costs no exploration resources. It does not replace any required business step.\n\nBusiness-work allowance:\n- backup work units: 126\n- charged public tool-call attempts: 40\n- write attempts: 20\nCharging rules:\n- Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope\'s refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity.\nExhaustion behavior: Further charged business operations are rejected when a required limit is exhausted.\nEpisode boundary: Committed protection state persists across episodes; episode limits reset.', 'business_budget': {'limits': [{'resource': 'backup work units', 'allowance': 126}, {'resource': 'charged public tool-call attempts', 'allowance': 40}, {'resource': 'write attempts', 'allowance': 20}], 'charging': ["Site-directory, exact workload, or exact policy reads cost 32 units; exact policy evidence costs 20 units, except a pending retained-snapshot refresh after observing a newer global epoch costs 14; ordinary repeated current queries still cost 20; each recovery dashboard invocation costs 7 and establishes the global epoch for all sites; repository listings cost 4 plus 2 per returned row; assignment reads cost 6 except exact current-episode own-assignment confirmation (site_id, policy_id, asset_id, limit=1, offset=0), which costs 0 work units and no call/write quota; schedule, activation, and assignment attempts cost 6, 6, and 8 units, including rejected writes. Observed global routing epochs and successfully refreshed exact policy evidence persist across tasks while that routing policy epoch remains in force. A task boundary alone does not invalidate them. A successful exact lookup reports evidence_validation for its returned scope; current evidence completes that scope's refresh without an assignment. Retained-snapshot evidence does not. Assignments still check current routing, repository profile, actor/site authority, policy/schedule state and available capacity."], 'exhaustion': 'Further charged business operations are rejected when a required limit is exhausted.', 'persistence': 'Committed protection state persists across episodes; episode limits reset.'}, 'completion': {'mode': 'independent', 'subset_allowed': True, 'priority_order': 'none', 'started_item_must_close': False, 'evidence': {'mode': 'narrow_status', 'source': 'list_policy_assignments with exact site_id, policy_id, asset_id, limit=1 and offset=0 after your own successful assignment in this episode', 'proves': "the submitted assignment's current active protection state"}}}] DATA_PROMPT = {'system_prompt': "You are operating in an unfamiliar persistent environment. Use only the available tools to complete the task within its stated limits. Do not claim that an action succeeded unless the tool confirms it.\n\nUse retained evidence to avoid repeating discovery of stable facts and rules. Distinguish reusable bindings from changing task data, availability, and other live state. Recheck when the operation requires fresh evidence, the relevant state may have changed, or current observations contradict what you retained. Current tool results take precedence. Estimate remaining work from the operations still needed, taking already known information into account.\n\nRecords from previous tasks may be available through the history tools. Consult them only when they help resolve a concrete uncertainty. The history tools share a limit of 30 calls per episode, so use them selectively. Previous records may be incomplete or outdated; verify relevant information against the current environment when necessary.\n\nCompletion evidence: a successful synchronous tool receipt is authoritative for the committed facts it explicitly reports. No additional readback is required merely to reconfirm those facts. A successful substep does not prove the whole artifact is complete. Pending, rejected, or failed operations are not completed operations. Perform any substantive validation, authorization, release, or other business step required by the task. If an operation requires a separate confirmation, use its declared status interface. Only explicitly designated narrow confirmation of your own current-task operations and quota readings are free; discovery and substantive business work retain their declared costs. Follow the task's completion and partial-delivery semantics and report only the outcomes established by current evidence.", 'task_prefix': 'Below is the task for this episode:\n\n', 'after_task_prompt': 'After completing or stopping this task, consider what the observed evidence could help resolve in future tasks in this environment.\n\nUse memory for concrete environment facts and executable bindings: retain the observed names or conditions, the exact public identifiers needed to act, their relationships, and relevant evidence or scope. Use skills for conditional rules and procedures: retain the observed conditions, valid or invalid actions and parameters, expected outcomes, and relevant costs. A generic instruction to look everything up again is not a substitute for the specific evidence learned.\n\nKeep notes compact and evidence-based. Preserve uncertainty when a relation has not been verified. Distinguish a cost that recurs for every new request from a discovery cost that retained knowledge can avoid. Update or remove contradicted entries rather than keeping incompatible claims. You decide which observations are worth retaining; if nothing useful was learned, do not update either store.', 'carrier_contract': 'Use memory for concrete environment facts and executable bindings: retain the observed names or conditions, the exact public identifiers needed to act, their relationships, and relevant evidence or scope. Use skills for conditional rules and procedures: retain the observed conditions, valid or invalid actions and parameters, expected outcomes, and relevant costs. A generic instruction to look everything up again is not a substitute for the specific evidence learned.', 'completion_evidence_profile': 'authoritative-receipt-v1'} DATA_SCHEDULE = {'schema': 'rescopebench-lifecycle-schedule/v1', 'task_count': 16, 'events': [{'event_id': 'E1', 'before_task': 5}]} DATA_POST_POLICY_CELLS = {('critical', 'restricted'): {'rule_atom': 'R01', 'service_tier': 'critical', 'data_class': 'restricted', 'protection_route': 'route-v2-a1', 'repository_profile': 'profile-delta'}, ('critical', 'confidential'): {'rule_atom': 'R02', 'service_tier': 'critical', 'data_class': 'confidential', 'protection_route': 'route-km4', 'repository_profile': 'profile-sigma'}, ('critical', 'internal'): {'rule_atom': 'R03', 'service_tier': 'critical', 'data_class': 'internal', 'protection_route': 'route-qn8', 'repository_profile': 'profile-ember'}, ('critical', 'public'): {'rule_atom': 'R04', 'service_tier': 'critical', 'data_class': 'public', 'protection_route': 'route-vp2', 'repository_profile': 'profile-delta'}, ('standard', 'restricted'): {'rule_atom': 'R05', 'service_tier': 'standard', 'data_class': 'restricted', 'protection_route': 'route-dc9', 'repository_profile': 'profile-delta'}, ('standard', 'confidential'): {'rule_atom': 'R06', 'service_tier': 'standard', 'data_class': 'confidential', 'protection_route': 'route-v2-b4', 'repository_profile': 'profile-ember'}, ('standard', 'internal'): {'rule_atom': 'R07', 'service_tier': 'standard', 'data_class': 'internal', 'protection_route': 'route-ls6', 'repository_profile': 'profile-sigma'}, ('standard', 'public'): {'rule_atom': 'R08', 'service_tier': 'standard', 'data_class': 'public', 'protection_route': 'route-rw5', 'repository_profile': 'profile-ember'}, ('archive', 'restricted'): {'rule_atom': 'R09', 'service_tier': 'archive', 'data_class': 'restricted', 'protection_route': 'route-bf1', 'repository_profile': 'profile-ember'}, ('archive', 'confidential'): {'rule_atom': 'R10', 'service_tier': 'archive', 'data_class': 'confidential', 'protection_route': 'route-gy8', 'repository_profile': 'profile-delta'}, ('archive', 'internal'): {'rule_atom': 'R11', 'service_tier': 'archive', 'data_class': 'internal', 'protection_route': 'route-v2-c7', 'repository_profile': 'profile-sigma'}, ('archive', 'public'): {'rule_atom': 'R12', 'service_tier': 'archive', 'data_class': 'public', 'protection_route': 'route-tz7', 'repository_profile': 'profile-sigma'}, ('regulated', 'restricted'): {'rule_atom': 'R13', 'service_tier': 'regulated', 'data_class': 'restricted', 'protection_route': 'route-cn2', 'repository_profile': 'profile-sigma'}, ('regulated', 'confidential'): {'rule_atom': 'R14', 'service_tier': 'regulated', 'data_class': 'confidential', 'protection_route': 'route-jp9', 'repository_profile': 'profile-ember'}, ('regulated', 'internal'): {'rule_atom': 'R15', 'service_tier': 'regulated', 'data_class': 'internal', 'protection_route': 'route-px5', 'repository_profile': 'profile-delta'}, ('regulated', 'public'): {'rule_atom': 'R16', 'service_tier': 'regulated', 'data_class': 'public', 'protection_route': 'route-v2-d9', 'repository_profile': 'profile-kappa'}} DATA_POLICY_CELLS = {('critical', 'restricted'): {'rule_atom': 'R01', 'service_tier': 'critical', 'data_class': 'restricted', 'protection_route': 'route-ax7', 'repository_profile': 'profile-kappa'}, ('critical', 'confidential'): {'rule_atom': 'R02', 'service_tier': 'critical', 'data_class': 'confidential', 'protection_route': 'route-km4', 'repository_profile': 'profile-sigma'}, ('critical', 'internal'): {'rule_atom': 'R03', 'service_tier': 'critical', 'data_class': 'internal', 'protection_route': 'route-qn8', 'repository_profile': 'profile-ember'}, ('critical', 'public'): {'rule_atom': 'R04', 'service_tier': 'critical', 'data_class': 'public', 'protection_route': 'route-vp2', 'repository_profile': 'profile-delta'}, ('standard', 'restricted'): {'rule_atom': 'R05', 'service_tier': 'standard', 'data_class': 'restricted', 'protection_route': 'route-dc9', 'repository_profile': 'profile-delta'}, ('standard', 'confidential'): {'rule_atom': 'R06', 'service_tier': 'standard', 'data_class': 'confidential', 'protection_route': 'route-hj3', 'repository_profile': 'profile-kappa'}, ('standard', 'internal'): {'rule_atom': 'R07', 'service_tier': 'standard', 'data_class': 'internal', 'protection_route': 'route-ls6', 'repository_profile': 'profile-sigma'}, ('standard', 'public'): {'rule_atom': 'R08', 'service_tier': 'standard', 'data_class': 'public', 'protection_route': 'route-rw5', 'repository_profile': 'profile-ember'}, ('archive', 'restricted'): {'rule_atom': 'R09', 'service_tier': 'archive', 'data_class': 'restricted', 'protection_route': 'route-bf1', 'repository_profile': 'profile-ember'}, ('archive', 'confidential'): {'rule_atom': 'R10', 'service_tier': 'archive', 'data_class': 'confidential', 'protection_route': 'route-gy8', 'repository_profile': 'profile-delta'}, ('archive', 'internal'): {'rule_atom': 'R11', 'service_tier': 'archive', 'data_class': 'internal', 'protection_route': 'route-mu4', 'repository_profile': 'profile-kappa'}, ('archive', 'public'): {'rule_atom': 'R12', 'service_tier': 'archive', 'data_class': 'public', 'protection_route': 'route-tz7', 'repository_profile': 'profile-sigma'}, ('regulated', 'restricted'): {'rule_atom': 'R13', 'service_tier': 'regulated', 'data_class': 'restricted', 'protection_route': 'route-cn2', 'repository_profile': 'profile-sigma'}, ('regulated', 'confidential'): {'rule_atom': 'R14', 'service_tier': 'regulated', 'data_class': 'confidential', 'protection_route': 'route-jp9', 'repository_profile': 'profile-ember'}, ('regulated', 'internal'): {'rule_atom': 'R15', 'service_tier': 'regulated', 'data_class': 'internal', 'protection_route': 'route-px5', 'repository_profile': 'profile-delta'}, ('regulated', 'public'): {'rule_atom': 'R16', 'service_tier': 'regulated', 'data_class': 'public', 'protection_route': 'route-wd3', 'repository_profile': 'profile-kappa'}} DATA_WRITE_TOOLS = {'create_backup_policy', 'assign_policy_to_asset', 'create_schedule_window', 'remove_policy_assignment', 'update_backup_policy_status'} DATA_PRE_RULES = {'R01': {'rule_atom': 'R01', 'service_tier': 'critical', 'data_class': 'restricted', 'protection_route': 'route-ax7', 'repository_profile': 'profile-kappa'}, 'R02': {'rule_atom': 'R02', 'service_tier': 'critical', 'data_class': 'confidential', 'protection_route': 'route-km4', 'repository_profile': 'profile-sigma'}, 'R03': {'rule_atom': 'R03', 'service_tier': 'critical', 'data_class': 'internal', 'protection_route': 'route-qn8', 'repository_profile': 'profile-ember'}, 'R04': {'rule_atom': 'R04', 'service_tier': 'critical', 'data_class': 'public', 'protection_route': 'route-vp2', 'repository_profile': 'profile-delta'}, 'R05': {'rule_atom': 'R05', 'service_tier': 'standard', 'data_class': 'restricted', 'protection_route': 'route-dc9', 'repository_profile': 'profile-delta'}, 'R06': {'rule_atom': 'R06', 'service_tier': 'standard', 'data_class': 'confidential', 'protection_route': 'route-hj3', 'repository_profile': 'profile-kappa'}, 'R07': {'rule_atom': 'R07', 'service_tier': 'standard', 'data_class': 'internal', 'protection_route': 'route-ls6', 'repository_profile': 'profile-sigma'}, 'R08': {'rule_atom': 'R08', 'service_tier': 'standard', 'data_class': 'public', 'protection_route': 'route-rw5', 'repository_profile': 'profile-ember'}, 'R09': {'rule_atom': 'R09', 'service_tier': 'archive', 'data_class': 'restricted', 'protection_route': 'route-bf1', 'repository_profile': 'profile-ember'}, 'R10': {'rule_atom': 'R10', 'service_tier': 'archive', 'data_class': 'confidential', 'protection_route': 'route-gy8', 'repository_profile': 'profile-delta'}, 'R11': {'rule_atom': 'R11', 'service_tier': 'archive', 'data_class': 'internal', 'protection_route': 'route-mu4', 'repository_profile': 'profile-kappa'}, 'R12': {'rule_atom': 'R12', 'service_tier': 'archive', 'data_class': 'public', 'protection_route': 'route-tz7', 'repository_profile': 'profile-sigma'}, 'R13': {'rule_atom': 'R13', 'service_tier': 'regulated', 'data_class': 'restricted', 'protection_route': 'route-cn2', 'repository_profile': 'profile-sigma'}, 'R14': {'rule_atom': 'R14', 'service_tier': 'regulated', 'data_class': 'confidential', 'protection_route': 'route-jp9', 'repository_profile': 'profile-ember'}, 'R15': {'rule_atom': 'R15', 'service_tier': 'regulated', 'data_class': 'internal', 'protection_route': 'route-px5', 'repository_profile': 'profile-delta'}, 'R16': {'rule_atom': 'R16', 'service_tier': 'regulated', 'data_class': 'public', 'protection_route': 'route-wd3', 'repository_profile': 'profile-kappa'}} DATA_TOOLS_2 = [{'name': 'list_protection_sites', 'description': 'List protection-site records visible to the current operator.', 'input_schema': {'type': 'object', 'properties': {'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False}}, {'name': 'get_site_recovery_dashboard', 'description': 'Return the bounded recovery summary for one site.', 'input_schema': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['site_id']}}, {'name': 'list_backup_repositories', 'description': 'List repository records for one site using optional filters.', 'input_schema': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'storage_tier_filter': {'type': 'string', 'minLength': 1}, 'status_filter': {'type': 'string', 'minLength': 1}, 'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False, 'required': ['site_id']}}, {'name': 'list_workload_assets', 'description': 'List workload-asset records for one site using optional filters.', 'input_schema': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'asset_type_filter': {'type': 'string', 'minLength': 1}, 'status_filter': {'type': 'string', 'minLength': 1}, 'classification_filter': {'type': 'string', 'minLength': 1}, 'environment_filter': {'type': 'string', 'minLength': 1}, 'tag_filter': {'type': 'string', 'minLength': 1}, 'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False, 'required': ['site_id']}}, {'name': 'get_workload_asset', 'description': 'Return one workload-asset record by asset ID.', 'input_schema': {'type': 'object', 'properties': {'asset_id': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['asset_id']}}, {'name': 'list_backup_policies', 'description': 'List backup-policy records for one site using optional filters.', 'input_schema': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'status_filter': {'type': 'string', 'minLength': 1}, 'policy_type_filter': {'type': 'string', 'minLength': 1}, 'replication_required': {'type': 'boolean'}, 'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False, 'required': ['site_id']}}, {'name': 'get_backup_policy', 'description': 'Return one backup-policy record by policy ID.', 'input_schema': {'type': 'object', 'properties': {'policy_id': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['policy_id']}}, {'name': 'list_policy_assignments', 'description': 'List policy-assignment records for one site using optional filters.', 'input_schema': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'policy_id': {'type': 'string', 'minLength': 1}, 'asset_id': {'type': 'string', 'minLength': 1}, 'state_filter': {'type': 'string', 'minLength': 1}, 'priority_filter': {'type': 'string', 'minLength': 1}, 'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False, 'required': ['site_id']}}, {'name': 'search_protection_policy_evidence', 'description': 'Search archived protection-policy evidence for one exact service-tier and data-class condition.', 'input_schema': {'type': 'object', 'properties': {'service_tier': {'type': 'string', 'minLength': 1}, 'data_class': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['service_tier', 'data_class']}}, {'name': 'create_backup_policy', 'description': 'Create one backup-policy record from the supplied policy fields.', 'input_schema': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'name': {'type': 'string', 'minLength': 1}, 'policy_type': {'type': 'string', 'minLength': 1}, 'backup_method': {'type': 'string', 'minLength': 1}, 'retention': {'type': 'object'}, 'encryption_required': {'type': 'boolean'}, 'immutable_required': {'type': 'boolean'}, 'replication_required': {'type': 'boolean'}, 'schedule_timezone': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['site_id', 'name', 'policy_type', 'backup_method', 'retention']}}, {'name': 'update_backup_policy_status', 'description': 'Update the lifecycle status of one backup-policy record.', 'input_schema': {'type': 'object', 'properties': {'policy_id': {'type': 'string', 'minLength': 1}, 'new_status': {'type': 'string', 'minLength': 1}, 'reason': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['policy_id', 'new_status']}}, {'name': 'create_schedule_window', 'description': 'Create one schedule-window record for a backup policy.', 'input_schema': {'type': 'object', 'properties': {'policy_id': {'type': 'string', 'minLength': 1}, 'name': {'type': 'string', 'minLength': 1}, 'frequency': {'type': 'string', 'minLength': 1}, 'start_time_local': {'type': 'string', 'minLength': 1}, 'duration_minutes': {'type': 'integer'}, 'allowed_weekdays': {'type': 'array', 'items': {'type': 'string', 'enum': ['mon', 'tue', 'wed', 'thu', 'fri', 'sat', 'sun']}, 'minItems': 1, 'uniqueItems': True}, 'blackout_dates': {'type': 'array', 'items': {'type': 'string', 'minLength': 1}, 'uniqueItems': True}, 'concurrency_limit': {'type': 'integer', 'minimum': 1, 'maximum': 50}, 'status': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['policy_id', 'name', 'frequency', 'start_time_local', 'duration_minutes', 'allowed_weekdays']}}, {'name': 'assign_policy_to_asset', 'description': 'Create one policy assignment for the supplied policy, asset, and repository IDs.', 'input_schema': {'type': 'object', 'properties': {'policy_id': {'type': 'string', 'minLength': 1}, 'asset_id': {'type': 'string', 'minLength': 1}, 'repository_id': {'type': 'string', 'minLength': 1}, 'protection_route': {'type': 'string', 'minLength': 1}, 'priority': {'type': 'string', 'minLength': 1}, 'protection_state': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['policy_id', 'asset_id', 'repository_id', 'protection_route']}}, {'name': 'remove_policy_assignment', 'description': 'Update one policy assignment to the removed state.', 'input_schema': {'type': 'object', 'properties': {'assignment_id': {'type': 'string', 'minLength': 1}, 'reason': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['assignment_id']}}] DATA_TASK_IDS = ('T01', 'T02', 'T03', 'T04', 'T05', 'T06', 'T07', 'T08', 'T09', 'T10', 'T11', 'T12', 'T13', 'T14', 'T15', 'T16') DATA_SCHEDULE_2 = {'schema': 'rescopebench-lifecycle-schedule/v1', 'task_count': 16, 'events': [{'event_id': 'E1', 'before_task': 5}]} DATA_CHANGED_RULE_IDS = ('R01', 'R06', 'R11', 'R16') DATA_ENTITY_ID_FIELDS = {'protection_site': 'site_id', 'operator_user': 'user_id', 'backup_repository': 'repository_id', 'workload_asset': 'asset_id', 'backup_policy': 'policy_id', 'schedule_window': 'schedule_id', 'policy_assignment': 'assignment_id', 'backup_job': 'job_id', 'restore_point': 'restore_point_id', 'restore_request': 'restore_request_id', 'replication_link': 'replication_link_id', 'replication_run': 'replication_run_id', 'recovery_test': 'test_id', 'compliance_report': 'report_id', 'failover_plan': 'failover_plan_id'} DATA_TOOL_BY_NAME = {'list_protection_sites': {'name': 'list_protection_sites', 'description': 'List protection-site records visible to the current operator.', 'input_schema': {'type': 'object', 'properties': {'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False}}, 'get_site_recovery_dashboard': {'name': 'get_site_recovery_dashboard', 'description': 'Return the bounded recovery summary for one site.', 'input_schema': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['site_id']}}, 'list_backup_repositories': {'name': 'list_backup_repositories', 'description': 'List repository records for one site using optional filters.', 'input_schema': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'storage_tier_filter': {'type': 'string', 'minLength': 1}, 'status_filter': {'type': 'string', 'minLength': 1}, 'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False, 'required': ['site_id']}}, 'list_workload_assets': {'name': 'list_workload_assets', 'description': 'List workload-asset records for one site using optional filters.', 'input_schema': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'asset_type_filter': {'type': 'string', 'minLength': 1}, 'status_filter': {'type': 'string', 'minLength': 1}, 'classification_filter': {'type': 'string', 'minLength': 1}, 'environment_filter': {'type': 'string', 'minLength': 1}, 'tag_filter': {'type': 'string', 'minLength': 1}, 'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False, 'required': ['site_id']}}, 'get_workload_asset': {'name': 'get_workload_asset', 'description': 'Return one workload-asset record by asset ID.', 'input_schema': {'type': 'object', 'properties': {'asset_id': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['asset_id']}}, 'list_backup_policies': {'name': 'list_backup_policies', 'description': 'List backup-policy records for one site using optional filters.', 'input_schema': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'status_filter': {'type': 'string', 'minLength': 1}, 'policy_type_filter': {'type': 'string', 'minLength': 1}, 'replication_required': {'type': 'boolean'}, 'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False, 'required': ['site_id']}}, 'get_backup_policy': {'name': 'get_backup_policy', 'description': 'Return one backup-policy record by policy ID.', 'input_schema': {'type': 'object', 'properties': {'policy_id': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['policy_id']}}, 'list_policy_assignments': {'name': 'list_policy_assignments', 'description': 'List policy-assignment records for one site using optional filters.', 'input_schema': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'policy_id': {'type': 'string', 'minLength': 1}, 'asset_id': {'type': 'string', 'minLength': 1}, 'state_filter': {'type': 'string', 'minLength': 1}, 'priority_filter': {'type': 'string', 'minLength': 1}, 'limit': {'type': 'integer', 'minimum': 1, 'maximum': 25}, 'offset': {'type': 'integer', 'minimum': 0}}, 'additionalProperties': False, 'required': ['site_id']}}, 'search_protection_policy_evidence': {'name': 'search_protection_policy_evidence', 'description': 'Search archived protection-policy evidence for one exact service-tier and data-class condition.', 'input_schema': {'type': 'object', 'properties': {'service_tier': {'type': 'string', 'minLength': 1}, 'data_class': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['service_tier', 'data_class']}}, 'create_backup_policy': {'name': 'create_backup_policy', 'description': 'Create one backup-policy record from the supplied policy fields.', 'input_schema': {'type': 'object', 'properties': {'site_id': {'type': 'string', 'minLength': 1}, 'name': {'type': 'string', 'minLength': 1}, 'policy_type': {'type': 'string', 'minLength': 1}, 'backup_method': {'type': 'string', 'minLength': 1}, 'retention': {'type': 'object'}, 'encryption_required': {'type': 'boolean'}, 'immutable_required': {'type': 'boolean'}, 'replication_required': {'type': 'boolean'}, 'schedule_timezone': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['site_id', 'name', 'policy_type', 'backup_method', 'retention']}}, 'update_backup_policy_status': {'name': 'update_backup_policy_status', 'description': 'Update the lifecycle status of one backup-policy record.', 'input_schema': {'type': 'object', 'properties': {'policy_id': {'type': 'string', 'minLength': 1}, 'new_status': {'type': 'string', 'minLength': 1}, 'reason': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['policy_id', 'new_status']}}, 'create_schedule_window': {'name': 'create_schedule_window', 'description': 'Create one schedule-window record for a backup policy.', 'input_schema': {'type': 'object', 'properties': {'policy_id': {'type': 'string', 'minLength': 1}, 'name': {'type': 'string', 'minLength': 1}, 'frequency': {'type': 'string', 'minLength': 1}, 'start_time_local': {'type': 'string', 'minLength': 1}, 'duration_minutes': {'type': 'integer'}, 'allowed_weekdays': {'type': 'array', 'items': {'type': 'string', 'enum': ['mon', 'tue', 'wed', 'thu', 'fri', 'sat', 'sun']}, 'minItems': 1, 'uniqueItems': True}, 'blackout_dates': {'type': 'array', 'items': {'type': 'string', 'minLength': 1}, 'uniqueItems': True}, 'concurrency_limit': {'type': 'integer', 'minimum': 1, 'maximum': 50}, 'status': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['policy_id', 'name', 'frequency', 'start_time_local', 'duration_minutes', 'allowed_weekdays']}}, 'assign_policy_to_asset': {'name': 'assign_policy_to_asset', 'description': 'Create one policy assignment for the supplied policy, asset, and repository IDs.', 'input_schema': {'type': 'object', 'properties': {'policy_id': {'type': 'string', 'minLength': 1}, 'asset_id': {'type': 'string', 'minLength': 1}, 'repository_id': {'type': 'string', 'minLength': 1}, 'protection_route': {'type': 'string', 'minLength': 1}, 'priority': {'type': 'string', 'minLength': 1}, 'protection_state': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['policy_id', 'asset_id', 'repository_id', 'protection_route']}}, 'remove_policy_assignment': {'name': 'remove_policy_assignment', 'description': 'Update one policy assignment to the removed state.', 'input_schema': {'type': 'object', 'properties': {'assignment_id': {'type': 'string', 'minLength': 1}, 'reason': {'type': 'string', 'minLength': 1}}, 'additionalProperties': False, 'required': ['assignment_id']}}} DATA_STRUCTURE_SCHEDULE = {'T01': ('S01', 'S01', 'S02', 'S03'), 'T02': ('S01', 'S01', 'S02', 'S03'), 'T03': ('S02', 'S03', 'S03', 'S04'), 'T04': ('S01', 'S02', 'S03', 'S04'), 'T05': ('S01', 'S02', 'S03', 'S04'), 'T06': ('S01', 'S02', 'S04', 'S03'), 'T07': ('S01', 'S02', 'S03', 'S04'), 'T08': ('S03', 'S01', 'S04', 'S02'), 'T09': ('S03', 'S01', 'S04', 'S02'), 'T10': ('S01', 'S02', 'S03', 'S04'), 'T11': ('S01', 'S02', 'S03', 'S04'), 'T12': ('S03', 'S04', 'S01', 'S02'), 'T13': ('S01', 'S02', 'S03', 'S04'), 'T14': ('S03', 'S04', 'S01', 'S02'), 'T15': ('S04', 'S01', 'S02', 'S03'), 'T16': ('S01', 'S03', 'S02', 'S04')} DATA_RULE_SCHEDULE = {'T01': ('R01', 'R06', 'R11', 'R16'), 'T02': ('R01', 'R06', 'R02', 'R07'), 'T03': ('R02', 'R07', 'R11', 'R16'), 'T04': ('R01', 'R06', 'R11', 'R16'), 'T05': ('R01', 'R06', 'R11', 'R16'), 'T06': ('R02', 'R07', 'R11', 'R16'), 'T07': ('R11', 'R16', 'R01', 'R06'), 'T08': ('R07', 'R01', 'R06', 'R02'), 'T09': ('R16', 'R11', 'R02', 'R07'), 'T10': ('R01', 'R07', 'R11', 'R06'), 'T11': ('R02', 'R06', 'R16', 'R01'), 'T12': ('R11', 'R16', 'R07', 'R02'), 'T13': ('R06', 'R02', 'R01', 'R16'), 'T14': ('R07', 'R11', 'R16', 'R06'), 'T15': ('R16', 'R01', 'R02', 'R07'), 'T16': ('R01', 'R11', 'R06', 'R16')} DATA_STRUCTURE_BY_ID = {'S01': {'atom_id': 'S01', 'site_id': 'site-dedb43cba4f0', 'site_name': 'Recovery Site A', 'repository_pool': 'pool-dedb43cba4f0', 'region': 'us-east'}, 'S02': {'atom_id': 'S02', 'site_id': 'site-2f5132072aa5', 'site_name': 'Recovery Site B', 'repository_pool': 'pool-2f5132072aa5', 'region': 'us-west'}, 'S03': {'atom_id': 'S03', 'site_id': 'site-cd843215d71e', 'site_name': 'Recovery Site C', 'repository_pool': 'pool-cd843215d71e', 'region': 'eu-central'}, 'S04': {'atom_id': 'S04', 'site_id': 'site-c54f3bbe9cee', 'site_name': 'Recovery Site D', 'repository_pool': 'pool-c54f3bbe9cee', 'region': 'ap-southeast'}, 'S05': {'atom_id': 'S05', 'site_id': 'site-cde04debd4f6', 'site_name': 'Recovery Site E', 'repository_pool': 'pool-cde04debd4f6', 'region': 'us-east'}, 'S06': {'atom_id': 'S06', 'site_id': 'site-cd1497178f8a', 'site_name': 'Recovery Site F', 'repository_pool': 'pool-cd1497178f8a', 'region': 'us-west'}, 'S07': {'atom_id': 'S07', 'site_id': 'site-064259a22d47', 'site_name': 'Recovery Site G', 'repository_pool': 'pool-064259a22d47', 'region': 'eu-central'}, 'S08': {'atom_id': 'S08', 'site_id': 'site-659579f6fbb3', 'site_name': 'Recovery Site H', 'repository_pool': 'pool-659579f6fbb3', 'region': 'ap-southeast'}} DATA_PROFILES = ('profile-kappa', 'profile-sigma', 'profile-ember', 'profile-delta') DATA_DRIFTED_POLICY_CELLS = {('critical', 'restricted'): {'rule_atom': 'R01', 'service_tier': 'critical', 'data_class': 'restricted', 'protection_route': 'route-v2-a1', 'repository_profile': 'profile-delta'}, ('critical', 'confidential'): {'rule_atom': 'R02', 'service_tier': 'critical', 'data_class': 'confidential', 'protection_route': 'route-km4', 'repository_profile': 'profile-sigma'}, ('critical', 'internal'): {'rule_atom': 'R03', 'service_tier': 'critical', 'data_class': 'internal', 'protection_route': 'route-qn8', 'repository_profile': 'profile-ember'}, ('critical', 'public'): {'rule_atom': 'R04', 'service_tier': 'critical', 'data_class': 'public', 'protection_route': 'route-vp2', 'repository_profile': 'profile-delta'}, ('standard', 'restricted'): {'rule_atom': 'R05', 'service_tier': 'standard', 'data_class': 'restricted', 'protection_route': 'route-dc9', 'repository_profile': 'profile-delta'}, ('standard', 'confidential'): {'rule_atom': 'R06', 'service_tier': 'standard', 'data_class': 'confidential', 'protection_route': 'route-v2-b4', 'repository_profile': 'profile-ember'}, ('standard', 'internal'): {'rule_atom': 'R07', 'service_tier': 'standard', 'data_class': 'internal', 'protection_route': 'route-ls6', 'repository_profile': 'profile-sigma'}, ('standard', 'public'): {'rule_atom': 'R08', 'service_tier': 'standard', 'data_class': 'public', 'protection_route': 'route-rw5', 'repository_profile': 'profile-ember'}, ('archive', 'restricted'): {'rule_atom': 'R09', 'service_tier': 'archive', 'data_class': 'restricted', 'protection_route': 'route-bf1', 'repository_profile': 'profile-ember'}, ('archive', 'confidential'): {'rule_atom': 'R10', 'service_tier': 'archive', 'data_class': 'confidential', 'protection_route': 'route-gy8', 'repository_profile': 'profile-delta'}, ('archive', 'internal'): {'rule_atom': 'R11', 'service_tier': 'archive', 'data_class': 'internal', 'protection_route': 'route-v2-c7', 'repository_profile': 'profile-sigma'}, ('archive', 'public'): {'rule_atom': 'R12', 'service_tier': 'archive', 'data_class': 'public', 'protection_route': 'route-tz7', 'repository_profile': 'profile-sigma'}, ('regulated', 'restricted'): {'rule_atom': 'R13', 'service_tier': 'regulated', 'data_class': 'restricted', 'protection_route': 'route-cn2', 'repository_profile': 'profile-sigma'}, ('regulated', 'confidential'): {'rule_atom': 'R14', 'service_tier': 'regulated', 'data_class': 'confidential', 'protection_route': 'route-jp9', 'repository_profile': 'profile-ember'}, ('regulated', 'internal'): {'rule_atom': 'R15', 'service_tier': 'regulated', 'data_class': 'internal', 'protection_route': 'route-px5', 'repository_profile': 'profile-delta'}, ('regulated', 'public'): {'rule_atom': 'R16', 'service_tier': 'regulated', 'data_class': 'public', 'protection_route': 'route-v2-d9', 'repository_profile': 'profile-kappa'}} DATA_POST_RULES = {'R01': {'rule_atom': 'R01', 'service_tier': 'critical', 'data_class': 'restricted', 'protection_route': 'route-v2-a1', 'repository_profile': 'profile-delta'}, 'R02': {'rule_atom': 'R02', 'service_tier': 'critical', 'data_class': 'confidential', 'protection_route': 'route-km4', 'repository_profile': 'profile-sigma'}, 'R03': {'rule_atom': 'R03', 'service_tier': 'critical', 'data_class': 'internal', 'protection_route': 'route-qn8', 'repository_profile': 'profile-ember'}, 'R04': {'rule_atom': 'R04', 'service_tier': 'critical', 'data_class': 'public', 'protection_route': 'route-vp2', 'repository_profile': 'profile-delta'}, 'R05': {'rule_atom': 'R05', 'service_tier': 'standard', 'data_class': 'restricted', 'protection_route': 'route-dc9', 'repository_profile': 'profile-delta'}, 'R06': {'rule_atom': 'R06', 'service_tier': 'standard', 'data_class': 'confidential', 'protection_route': 'route-v2-b4', 'repository_profile': 'profile-ember'}, 'R07': {'rule_atom': 'R07', 'service_tier': 'standard', 'data_class': 'internal', 'protection_route': 'route-ls6', 'repository_profile': 'profile-sigma'}, 'R08': {'rule_atom': 'R08', 'service_tier': 'standard', 'data_class': 'public', 'protection_route': 'route-rw5', 'repository_profile': 'profile-ember'}, 'R09': {'rule_atom': 'R09', 'service_tier': 'archive', 'data_class': 'restricted', 'protection_route': 'route-bf1', 'repository_profile': 'profile-ember'}, 'R10': {'rule_atom': 'R10', 'service_tier': 'archive', 'data_class': 'confidential', 'protection_route': 'route-gy8', 'repository_profile': 'profile-delta'}, 'R11': {'rule_atom': 'R11', 'service_tier': 'archive', 'data_class': 'internal', 'protection_route': 'route-v2-c7', 'repository_profile': 'profile-sigma'}, 'R12': {'rule_atom': 'R12', 'service_tier': 'archive', 'data_class': 'public', 'protection_route': 'route-tz7', 'repository_profile': 'profile-sigma'}, 'R13': {'rule_atom': 'R13', 'service_tier': 'regulated', 'data_class': 'restricted', 'protection_route': 'route-cn2', 'repository_profile': 'profile-sigma'}, 'R14': {'rule_atom': 'R14', 'service_tier': 'regulated', 'data_class': 'confidential', 'protection_route': 'route-jp9', 'repository_profile': 'profile-ember'}, 'R15': {'rule_atom': 'R15', 'service_tier': 'regulated', 'data_class': 'internal', 'protection_route': 'route-px5', 'repository_profile': 'profile-delta'}, 'R16': {'rule_atom': 'R16', 'service_tier': 'regulated', 'data_class': 'public', 'protection_route': 'route-v2-d9', 'repository_profile': 'profile-kappa'}} DATA_LIST_KEYS = {'list_protection_sites': 'sites', 'list_backup_repositories': 'repositories', 'list_workload_assets': 'assets', 'list_backup_policies': 'policies'} DATA_FIXED_COSTS = {'get_site_recovery_dashboard': 7, 'get_workload_asset': 7, 'get_backup_policy': 7, 'list_policy_assignments': 6, 'search_protection_policy_evidence': 16, 'create_backup_policy': 10, 'update_backup_policy_status': 6, 'create_schedule_window': 6, 'assign_policy_to_asset': 8, 'remove_policy_assignment': 8} DATA_RULE_SCHEDULE_2 = {'T01': ('R01', 'R06', 'R11', 'R16'), 'T02': ('R01', 'R06', 'R11', 'R16'), 'T03': ('R01', 'R06', 'R01', 'R11'), 'T04': ('R01', 'R06', 'R11', 'R16'), 'T05': ('R01', 'R06', 'R11', 'R16'), 'T06': ('R01', 'R06', 'R11', 'R16'), 'T07': ('R01', 'R06', 'R16', 'R11'), 'T08': ('R11', 'R01', 'R16', 'R06'), 'T09': ('R11', 'R01', 'R16', 'R06'), 'T10': ('R01', 'R06', 'R11', 'R16'), 'T11': ('R01', 'R06', 'R11', 'R16'), 'T12': ('R11', 'R16', 'R01', 'R06'), 'T13': ('R01', 'R06', 'R11', 'R16'), 'T14': ('R11', 'R16', 'R01', 'R06'), 'T15': ('R16', 'R01', 'R06', 'R11'), 'T16': ('R01', 'R11', 'R06', 'R16')} DATA_STRUCTURE_SCHEDULE_2 = {'T01': ('S01', 'S01', 'S02', 'S03'), 'T02': ('S01', 'S01', 'S02', 'S03'), 'T03': ('S01', 'S01', 'S02', 'S03'), 'T04': ('S01', 'S02', 'S01', 'S03'), 'T05': ('S01', 'S02', 'S03', 'S04'), 'T06': ('S01', 'S02', 'S04', 'S03'), 'T07': ('S01', 'S02', 'S03', 'S04'), 'T08': ('S03', 'S01', 'S04', 'S02'), 'T09': ('S03', 'S01', 'S04', 'S02'), 'T10': ('S01', 'S02', 'S03', 'S04'), 'T11': ('S01', 'S02', 'S03', 'S04'), 'T12': ('S03', 'S04', 'S01', 'S02'), 'T13': ('S01', 'S02', 'S03', 'S04'), 'T14': ('S03', 'S04', 'S01', 'S02'), 'T15': ('S04', 'S01', 'S02', 'S03'), 'T16': ('S01', 'S03', 'S02', 'S04')} DATA_COMPLETION = {'mode': 'ordered_independent', 'subset_allowed': True, 'priority_order': 'listed', 'started_item_must_close': True, 'verification_required': True} DATA_RULE_SCHEDULE_3 = {'T01': ('R01', 'R06', 'R11', 'R16'), 'T02': ('R01', 'R06', 'R11', 'R16'), 'T03': ('R01', 'R06', 'R01', 'R11'), 'T04': ('R01', 'R06', 'R11', 'R16'), 'T05': ('R01', 'R06', 'R11', 'R16'), 'T06': ('R01', 'R06', 'R11', 'R16'), 'T07': ('R01', 'R06', 'R16', 'R11'), 'T08': ('R11', 'R01', 'R16', 'R06'), 'T09': ('R11', 'R01', 'R16', 'R06'), 'T10': ('R01', 'R06', 'R11', 'R16'), 'T11': ('R01', 'R06', 'R11', 'R16'), 'T12': ('R16', 'R01', 'R06', 'R11'), 'T13': ('R01', 'R06', 'R11', 'R16'), 'T14': ('R11', 'R16', 'R01', 'R06'), 'T15': ('R16', 'R01', 'R06', 'R11'), 'T16': ('R01', 'R11', 'R06', 'R16')} DATA_STRUCTURE_SCHEDULE_3 = {'T01': ('S01', 'S01', 'S02', 'S03'), 'T02': ('S01', 'S01', 'S02', 'S03'), 'T03': ('S01', 'S01', 'S02', 'S03'), 'T04': ('S01', 'S02', 'S01', 'S03'), 'T05': ('S01', 'S02', 'S03', 'S04'), 'T06': ('S01', 'S02', 'S04', 'S03'), 'T07': ('S01', 'S02', 'S03', 'S04'), 'T08': ('S03', 'S01', 'S04', 'S02'), 'T09': ('S03', 'S01', 'S04', 'S02'), 'T10': ('S01', 'S02', 'S03', 'S04'), 'T11': ('S01', 'S02', 'S03', 'S04'), 'T12': ('S04', 'S01', 'S02', 'S03'), 'T13': ('S01', 'S02', 'S03', 'S04'), 'T14': ('S03', 'S04', 'S01', 'S02'), 'T15': ('S04', 'S01', 'S02', 'S03'), 'T16': ('S01', 'S03', 'S02', 'S04')} DATA_STRUCTURE_CLUSTERS = ({'atom_id': 'S01', 'site_id': 'site-dedb43cba4f0', 'site_name': 'Recovery Site A', 'repository_pool': 'pool-dedb43cba4f0', 'region': 'us-east'}, {'atom_id': 'S02', 'site_id': 'site-2f5132072aa5', 'site_name': 'Recovery Site B', 'repository_pool': 'pool-2f5132072aa5', 'region': 'us-west'}, {'atom_id': 'S03', 'site_id': 'site-cd843215d71e', 'site_name': 'Recovery Site C', 'repository_pool': 'pool-cd843215d71e', 'region': 'eu-central'}, {'atom_id': 'S04', 'site_id': 'site-c54f3bbe9cee', 'site_name': 'Recovery Site D', 'repository_pool': 'pool-c54f3bbe9cee', 'region': 'ap-southeast'}, {'atom_id': 'S05', 'site_id': 'site-cde04debd4f6', 'site_name': 'Recovery Site E', 'repository_pool': 'pool-cde04debd4f6', 'region': 'us-east'}, {'atom_id': 'S06', 'site_id': 'site-cd1497178f8a', 'site_name': 'Recovery Site F', 'repository_pool': 'pool-cd1497178f8a', 'region': 'us-west'}, {'atom_id': 'S07', 'site_id': 'site-064259a22d47', 'site_name': 'Recovery Site G', 'repository_pool': 'pool-064259a22d47', 'region': 'eu-central'}, {'atom_id': 'S08', 'site_id': 'site-659579f6fbb3', 'site_name': 'Recovery Site H', 'repository_pool': 'pool-659579f6fbb3', 'region': 'ap-southeast'}) DATA_RULE_SCHEDULE_4 = {'T01': ('R01', 'R06', 'R11', 'R16'), 'T02': ('R02', 'R07', 'R12', 'R13'), 'T03': ('R03', 'R08', 'R09', 'R14'), 'T04': ('R04', 'R05', 'R10', 'R15'), 'T05': ('R01', 'R07', 'R11', 'R13'), 'T06': ('R02', 'R06', 'R12', 'R16'), 'T07': ('R03', 'R05', 'R09', 'R15'), 'T08': ('R01', 'R02', 'R06', 'R07'), 'T09': ('R01', 'R02', 'R06', 'R07'), 'T10': ('R11', 'R12', 'R16', 'R15'), 'T11': ('R01', 'R03', 'R06', 'R08'), 'T12': ('R11', 'R04', 'R16', 'R05'), 'T13': ('R01', 'R09', 'R06', 'R14'), 'T14': ('R11', 'R02', 'R16', 'R07'), 'T15': ('R01', 'R12', 'R06', 'R13'), 'T16': ('R11', 'R03', 'R16', 'R08')} DATA_STRUCTURE_SCHEDULE_4 = {'T01': ('S01', 'S02', 'S01', 'S02'), 'T02': ('S03', 'S04', 'S03', 'S04'), 'T03': ('S05', 'S06', 'S05', 'S06'), 'T04': ('S07', 'S08', 'S07', 'S08'), 'T05': ('S01', 'S03', 'S05', 'S07'), 'T06': ('S02', 'S04', 'S06', 'S08'), 'T07': ('S01', 'S04', 'S05', 'S08'), 'T08': ('S02', 'S03', 'S06', 'S07'), 'T09': ('S02', 'S03', 'S06', 'S07'), 'T10': ('S01', 'S04', 'S05', 'S08'), 'T11': ('S01', 'S03', 'S06', 'S08'), 'T12': ('S02', 'S04', 'S05', 'S07'), 'T13': ('S01', 'S04', 'S06', 'S07'), 'T14': ('S02', 'S03', 'S05', 'S08'), 'T15': ('S01', 'S02', 'S07', 'S08'), 'T16': ('S03', 'S04', 'S05', 'S06')}