Tool tree
Choose a reference architecture, stress the live bench, and watch human decisions and WebMCP remediation meet in one state.
Implementation
gradeNot graded
gradeNot graded
page: Every page
get_webmcp_statusAnswer
Read WebMCP capability, toolsReady, session id, and registered tool names. Call this before other tools if discovery looks incomplete.
View tool JSON
{
"name": "get_webmcp_status",
"kind": "answer",
"impl": "imperative",
"description": "Read WebMCP capability, toolsReady, session id, and registered tool names. Call this before other tools if discovery looks incomplete.",
"inputSchema": {
"type": "object",
"properties": {},
"required": [],
"additionalProperties": false
},
"page": "/"
}get_catalogue_guideAnswer
Read how to pick a catalogue simulation and load it onto the live bench.
View tool JSON
{
"name": "get_catalogue_guide",
"kind": "answer",
"impl": "imperative",
"description": "Read how to pick a catalogue simulation and load it onto the live bench.",
"inputSchema": {
"type": "object",
"properties": {},
"required": [],
"additionalProperties": false
},
"page": "/"
}list_architecturesAnswer
List the three equal GCP reference simulations available to load onto the bench.
View tool JSON
{
"name": "list_architectures",
"kind": "answer",
"impl": "imperative",
"description": "List the three equal GCP reference simulations available to load onto the bench.",
"inputSchema": {
"type": "object",
"properties": {},
"required": [],
"additionalProperties": false
},
"page": "/"
}load_architectureAction
Load a catalogue simulation onto the live bench. Accepts architecture id (event_driven_checkout, multi_region_saas, llm_inference_serving) or a short name such as checkout, saas, or llm. Navigates to /bench/:id.
View tool JSON
{
"name": "load_architecture",
"kind": "act",
"impl": "imperative",
"description": "Load a catalogue simulation onto the live bench. Accepts architecture id (event_driven_checkout, multi_region_saas, llm_inference_serving) or a short name such as checkout, saas, or llm. Navigates to /bench/:id.",
"inputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string",
"description": "Architecture id or short name: checkout, saas, llm, or the full id."
}
},
"required": [
"id"
],
"additionalProperties": false
},
"page": "/"
}get_bench_guideAnswer
Read the signature human-agent loop, valid remediation paths, pin semantics, and error codes for this bench.
View tool JSON
{
"name": "get_bench_guide",
"kind": "answer",
"impl": "imperative",
"description": "Read the signature human-agent loop, valid remediation paths, pin semantics, and error codes for this bench.",
"inputSchema": {
"type": "object",
"properties": {},
"required": [],
"additionalProperties": false
},
"page": "/"
}get_architectureAnswer
Read the current semantic topology, health, failures, exclusions, and store version.
View tool JSON
{
"name": "get_architecture",
"kind": "answer",
"impl": "imperative",
"description": "Read the current semantic topology, health, failures, exclusions, and store version.",
"inputSchema": {
"type": "object",
"properties": {},
"required": [],
"additionalProperties": false
},
"page": "/"
}get_scenarioAnswer
Read current human scenario controls, targets, pins, and store version.
View tool JSON
{
"name": "get_scenario",
"kind": "answer",
"impl": "imperative",
"description": "Read current human scenario controls, targets, pins, and store version.",
"inputSchema": {
"type": "object",
"properties": {},
"required": [],
"additionalProperties": false
},
"page": "/"
}get_live_metricsAnswer
Read demand, served throughput, capacity, utilisation, traffic share, measurement semantics, traffic-weighted fault contributions, and latency-routing scores with intended versus effective split.
View tool JSON
{
"name": "get_live_metrics",
"kind": "answer",
"impl": "imperative",
"description": "Read demand, served throughput, capacity, utilisation, traffic share, measurement semantics, traffic-weighted fault contributions, and latency-routing scores with intended versus effective split.",
"inputSchema": {
"type": "object",
"properties": {},
"required": [],
"additionalProperties": false
},
"page": "/"
}get_root_cause_analysisAnswer
Explain the current failure using live topology, fault, capacity, SLO, and Flight Data Recorder evidence. Pass unavailableTargets and prohibitedActions when a recovery path is blocked. Returns recommendedActions plus a ready-to-preview remediationPlan when a compound surviving-path recovery is legal.
View tool JSON
{
"name": "get_root_cause_analysis",
"kind": "answer",
"impl": "imperative",
"description": "Explain the current failure using live topology, fault, capacity, SLO, and Flight Data Recorder evidence. Pass unavailableTargets and prohibitedActions when a recovery path is blocked. Returns recommendedActions plus a ready-to-preview remediationPlan when a compound surviving-path recovery is legal.",
"inputSchema": {
"type": "object",
"properties": {
"unavailableTargets": {
"type": "array",
"items": {
"type": "string"
},
"description": "Regions, nodes, or edges that cannot be restored or cleared, such as europe-west2."
},
"prohibitedActions": {
"type": "array",
"items": {
"type": "string"
},
"description": "Tool names that must not be recommended, such as clear_fault_profile."
}
},
"required": [],
"additionalProperties": false
},
"page": "/"
}get_constraintsAnswer
Read applicable provider-limit and model-assumption records with source dates and provenance.
View tool JSON
{
"name": "get_constraints",
"kind": "answer",
"impl": "imperative",
"description": "Read applicable provider-limit and model-assumption records with source dates and provenance.",
"inputSchema": {
"type": "object",
"properties": {},
"required": [],
"additionalProperties": false
},
"page": "/"
}get_bench_snapshotAnswer
Atomically read scenario, topology, metrics, constraints, RCA, version, tick, and timestamp from one simulation tick.
View tool JSON
{
"name": "get_bench_snapshot",
"kind": "answer",
"impl": "imperative",
"description": "Atomically read scenario, topology, metrics, constraints, RCA, version, tick, and timestamp from one simulation tick.",
"inputSchema": {
"type": "object",
"properties": {},
"required": [],
"additionalProperties": false
},
"page": "/"
}get_decision_logAnswer
Read the Flight Data Recorder: recent ui, webmcp, and sim operator events with versions and result codes. Use after STALE_STATE to see why the store version moved. Does not increment storeVersion.
View tool JSON
{
"name": "get_decision_log",
"kind": "answer",
"impl": "imperative",
"description": "Read the Flight Data Recorder: recent ui, webmcp, and sim operator events with versions and result codes. Use after STALE_STATE to see why the store version moved. Does not increment storeVersion.",
"inputSchema": {
"type": "object",
"properties": {},
"required": [],
"additionalProperties": false
},
"page": "/"
}preview_changeAction
Read-only projection of a mutation. Does not increment storeVersion. projectionSucceeded is true when the nested op would apply; it is never an applied write.
View tool JSON
{
"name": "preview_change",
"kind": "act",
"impl": "imperative",
"description": "Read-only projection of a mutation. Does not increment storeVersion. projectionSucceeded is true when the nested op would apply; it is never an applied write.",
"inputSchema": {
"type": "object",
"properties": {
"op": {
"type": "string"
},
"args": {
"type": "object"
}
},
"required": [
"op"
],
"additionalProperties": false
},
"page": "/"
}apply_remediation_planAction
Apply several legal mutations as one all-or-nothing transaction with a single expectedVersion and one version bump.
View tool JSON
{
"name": "apply_remediation_plan",
"kind": "act",
"impl": "imperative",
"description": "Apply several legal mutations as one all-or-nothing transaction with a single expectedVersion and one version bump.",
"inputSchema": {
"type": "object",
"properties": {
"steps": {
"type": "array",
"minItems": 1,
"maxItems": 8,
"items": {
"type": "object",
"properties": {
"op": {
"type": "string"
},
"args": {
"type": "object"
}
},
"required": [
"op"
],
"additionalProperties": false
}
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"steps",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}set_peak_rpsAction
Set peak load demand in requests per second for the current scenario.
View tool JSON
{
"name": "set_peak_rps",
"kind": "act",
"impl": "imperative",
"description": "Set peak load demand in requests per second for the current scenario.",
"inputSchema": {
"type": "object",
"properties": {
"peakRps": {
"type": "number"
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"peakRps",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}set_budgetAction
Set the monthly budget constraint in GBP for the current scenario.
View tool JSON
{
"name": "set_budget",
"kind": "act",
"impl": "imperative",
"description": "Set the monthly budget constraint in GBP for the current scenario.",
"inputSchema": {
"type": "object",
"properties": {
"budget": {
"type": "number"
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"budget",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}run_stress_testAction
Start the deterministic stress or failure path for the loaded reference.
View tool JSON
{
"name": "run_stress_test",
"kind": "act",
"impl": "imperative",
"description": "Start the deterministic stress or failure path for the loaded reference.",
"inputSchema": {
"type": "object",
"properties": {
"expectedVersion": {
"type": "number"
}
},
"required": [
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}stop_stress_testAction
Stop the running stress test and leave measured failure state. Use reset_scenario to return the visible bench to baseline.
View tool JSON
{
"name": "stop_stress_test",
"kind": "act",
"impl": "imperative",
"description": "Stop the running stress test and leave measured failure state. Use reset_scenario to return the visible bench to baseline.",
"inputSchema": {
"type": "object",
"properties": {
"expectedVersion": {
"type": "number"
}
},
"required": [
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}reset_scenarioAction
Reset the loaded reference to its operational baseline: tick 0, stress off, no faults or outages, default load and traffic split. Human pins are kept. Updates the visible topology and gauges.
View tool JSON
{
"name": "reset_scenario",
"kind": "act",
"impl": "imperative",
"description": "Reset the loaded reference to its operational baseline: tick 0, stress off, no faults or outages, default load and traffic split. Human pins are kept. Updates the visible topology and gauges.",
"inputSchema": {
"type": "object",
"properties": {
"expectedVersion": {
"type": "number"
}
},
"required": [
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}fail_componentAction
Fail a runtime component as a hard outage. Distinct from set_fault_profile dropout, which is packet loss.
View tool JSON
{
"name": "fail_component",
"kind": "act",
"impl": "imperative",
"description": "Fail a runtime component as a hard outage. Distinct from set_fault_profile dropout, which is packet loss.",
"inputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string",
"enum": [
"web_client",
"api_gateway",
"cloud_run_order",
"pubsub_ordered",
"cloud_run_payment",
"cloud_sql",
"memorystore",
"global_alb",
"cloud_run_europe",
"cloud_run_us",
"app_europe",
"app_us",
"cloud_sql_primary",
"cloud_sql_replica",
"clients",
"cloud_run_router",
"vertex_stable",
"vertex_rc",
"pubsub_overflow"
]
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"id",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}restore_componentAction
Restore a runtime component that was killed or failed.
View tool JSON
{
"name": "restore_component",
"kind": "act",
"impl": "imperative",
"description": "Restore a runtime component that was killed or failed.",
"inputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string",
"enum": [
"web_client",
"api_gateway",
"cloud_run_order",
"pubsub_ordered",
"cloud_run_payment",
"cloud_sql",
"memorystore",
"global_alb",
"cloud_run_europe",
"cloud_run_us",
"app_europe",
"app_us",
"cloud_sql_primary",
"cloud_sql_replica",
"clients",
"cloud_run_router",
"vertex_stable",
"vertex_rc",
"pubsub_overflow"
]
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"id",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}fail_zoneAction
Fail one configured availability zone while leaving the topology and surviving replicas visible.
View tool JSON
{
"name": "fail_zone",
"kind": "act",
"impl": "imperative",
"description": "Fail one configured availability zone while leaving the topology and surviving replicas visible.",
"inputSchema": {
"type": "object",
"properties": {
"zone": {
"type": "string",
"enum": [
"europe-west2-a",
"europe-west2-b",
"us-east4-a",
"us-east4-b"
]
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"zone",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}restore_zoneAction
Restore one failed availability zone and its placed replicas.
View tool JSON
{
"name": "restore_zone",
"kind": "act",
"impl": "imperative",
"description": "Restore one failed availability zone and its placed replicas.",
"inputSchema": {
"type": "object",
"properties": {
"zone": {
"type": "string",
"enum": [
"europe-west2-a",
"europe-west2-b",
"us-east4-a",
"us-east4-b"
]
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"zone",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}set_fault_profileAction
Inject deterministic latency (ms) and packet loss (%) at any declared component or connection. packetLossPercent is the unit-explicit name; dropoutPercent is accepted as an alias.
View tool JSON
{
"name": "set_fault_profile",
"kind": "act",
"impl": "imperative",
"description": "Inject deterministic latency (ms) and packet loss (%) at any declared component or connection. packetLossPercent is the unit-explicit name; dropoutPercent is accepted as an alias.",
"inputSchema": {
"type": "object",
"properties": {
"targetId": {
"type": "string",
"enum": [
"web_client",
"api_gateway",
"cloud_run_order",
"pubsub_ordered",
"cloud_run_payment",
"cloud_sql",
"memorystore",
"checkout-client-api",
"checkout-api-order",
"checkout-order-pubsub",
"checkout-pubsub-payment",
"checkout-order-db",
"checkout-order-cache",
"global_alb",
"cloud_run_europe",
"cloud_run_us",
"app_europe",
"app_us",
"cloud_sql_primary",
"cloud_sql_replica",
"region-alb-europe",
"region-alb-us",
"region-ingress-app-europe",
"region-ingress-app-us",
"region-app-sql-europe",
"region-app-sql-us",
"region-sql-cache-europe",
"region-sql-cache-us",
"clients",
"cloud_run_router",
"vertex_stable",
"vertex_rc",
"pubsub_overflow",
"llm-client-api",
"llm-api-router",
"llm-router-stable",
"llm-router-rc",
"llm-stable-cache",
"llm-rc-overflow"
]
},
"latencyMs": {
"type": "number",
"minimum": 0,
"maximum": 30000,
"description": "Added latency in milliseconds."
},
"packetLossPercent": {
"type": "number",
"minimum": 0,
"maximum": 100,
"description": "Packet loss as a percent of requests (0-100)."
},
"dropoutPercent": {
"type": "number",
"minimum": 0,
"maximum": 100,
"description": "Deprecated alias of packetLossPercent."
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"targetId",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}set_region_fault_profileAction
Apply the same latency (ms) and packet loss (%) to a regional boundary. Accepts europe-west2 / West2 / us-east4 / east4. On multi-region SaaS this is the ALB ingress edge, not a guessed node id.
View tool JSON
{
"name": "set_region_fault_profile",
"kind": "act",
"impl": "imperative",
"description": "Apply the same latency (ms) and packet loss (%) to a regional boundary. Accepts europe-west2 / West2 / us-east4 / east4. On multi-region SaaS this is the ALB ingress edge, not a guessed node id.",
"inputSchema": {
"type": "object",
"properties": {
"region": {
"type": "string",
"description": "Region id or alias such as europe-west2, West2, us-east4, east4."
},
"latencyMs": {
"type": "number",
"minimum": 0,
"maximum": 30000,
"description": "Added latency in milliseconds."
},
"packetLossPercent": {
"type": "number",
"minimum": 0,
"maximum": 100,
"description": "Packet loss as a percent of requests (0-100)."
},
"dropoutPercent": {
"type": "number",
"minimum": 0,
"maximum": 100,
"description": "Deprecated alias of packetLossPercent."
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"region",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}ramp_fault_untilAction
Ramp latencyMs or packetLossPercent from start toward ceiling in steps. Records every intermediate result and stops when the stop condition matches. One version bump. Does not require the agent to issue N stale-checked mutations.
View tool JSON
{
"name": "ramp_fault_until",
"kind": "act",
"impl": "imperative",
"description": "Ramp latencyMs or packetLossPercent from start toward ceiling in steps. Records every intermediate result and stops when the stop condition matches. One version bump. Does not require the agent to issue N stale-checked mutations.",
"inputSchema": {
"type": "object",
"properties": {
"targetId": {
"type": "string",
"enum": [
"web_client",
"api_gateway",
"cloud_run_order",
"pubsub_ordered",
"cloud_run_payment",
"cloud_sql",
"memorystore",
"checkout-client-api",
"checkout-api-order",
"checkout-order-pubsub",
"checkout-pubsub-payment",
"checkout-order-db",
"checkout-order-cache",
"global_alb",
"cloud_run_europe",
"cloud_run_us",
"app_europe",
"app_us",
"cloud_sql_primary",
"cloud_sql_replica",
"region-alb-europe",
"region-alb-us",
"region-ingress-app-europe",
"region-ingress-app-us",
"region-app-sql-europe",
"region-app-sql-us",
"region-sql-cache-europe",
"region-sql-cache-us",
"clients",
"cloud_run_router",
"vertex_stable",
"vertex_rc",
"pubsub_overflow",
"llm-client-api",
"llm-api-router",
"llm-router-stable",
"llm-router-rc",
"llm-stable-cache",
"llm-rc-overflow"
]
},
"region": {
"type": "string",
"description": "Optional region alias. When set, ramps set_region_fault_profile instead of a single target."
},
"metric": {
"type": "string",
"enum": [
"latencyMs",
"packetLossPercent"
]
},
"start": {
"type": "number"
},
"step": {
"type": "number"
},
"ceiling": {
"type": "number"
},
"stop": {
"type": "object",
"properties": {
"metric": {
"type": "string",
"enum": [
"availability",
"p95Ms",
"errorRate",
"sloStatus",
"effectivePrimaryPercent"
]
},
"op": {
"type": "string",
"enum": [
"lt",
"lte",
"gt",
"gte",
"eq"
]
},
"value": {
"type": [
"number",
"string"
]
}
},
"required": [
"metric",
"op",
"value"
],
"additionalProperties": false
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"metric",
"start",
"step",
"ceiling",
"stop",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}clear_fault_profileAction
Clear injected latency and packet loss from one component or connection.
View tool JSON
{
"name": "clear_fault_profile",
"kind": "act",
"impl": "imperative",
"description": "Clear injected latency and packet loss from one component or connection.",
"inputSchema": {
"type": "object",
"properties": {
"targetId": {
"type": "string",
"enum": [
"web_client",
"api_gateway",
"cloud_run_order",
"pubsub_ordered",
"cloud_run_payment",
"cloud_sql",
"memorystore",
"checkout-client-api",
"checkout-api-order",
"checkout-order-pubsub",
"checkout-pubsub-payment",
"checkout-order-db",
"checkout-order-cache",
"global_alb",
"cloud_run_europe",
"cloud_run_us",
"app_europe",
"app_us",
"cloud_sql_primary",
"cloud_sql_replica",
"region-alb-europe",
"region-alb-us",
"region-ingress-app-europe",
"region-ingress-app-us",
"region-app-sql-europe",
"region-app-sql-us",
"region-sql-cache-europe",
"region-sql-cache-us",
"clients",
"cloud_run_router",
"vertex_stable",
"vertex_rc",
"pubsub_overflow",
"llm-client-api",
"llm-api-router",
"llm-router-stable",
"llm-router-rc",
"llm-stable-cache",
"llm-rc-overflow"
]
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"targetId",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}clear_all_faultsAction
Clear every injected latency and packet-loss profile while preserving load, scaling, traffic split, pins, and zone state.
View tool JSON
{
"name": "clear_all_faults",
"kind": "act",
"impl": "imperative",
"description": "Clear every injected latency and packet-loss profile while preserving load, scaling, traffic split, pins, and zone state.",
"inputSchema": {
"type": "object",
"properties": {
"expectedVersion": {
"type": "number"
}
},
"required": [
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}set_autoscalingAction
Set autoscaling bounds. Replicas follow demand toward the target utilisation within min/max. When a zone is failed, replacement replicas are placed in surviving zones of the same region.
View tool JSON
{
"name": "set_autoscaling",
"kind": "act",
"impl": "imperative",
"description": "Set autoscaling bounds. Replicas follow demand toward the target utilisation within min/max. When a zone is failed, replacement replicas are placed in surviving zones of the same region.",
"inputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string",
"enum": [
"web_client",
"api_gateway",
"cloud_run_order",
"pubsub_ordered",
"cloud_run_payment",
"cloud_sql",
"memorystore",
"global_alb",
"cloud_run_europe",
"cloud_run_us",
"app_europe",
"app_us",
"cloud_sql_primary",
"cloud_sql_replica",
"clients",
"cloud_run_router",
"vertex_stable",
"vertex_rc",
"pubsub_overflow"
]
},
"min": {
"type": "number",
"minimum": 1,
"maximum": 32
},
"max": {
"type": "number",
"minimum": 1,
"maximum": 32
},
"targetUtilPercent": {
"type": "number",
"minimum": 10,
"maximum": 90
},
"enabled": {
"type": "boolean"
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"id",
"min",
"max",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}set_ordering_key_parallelismAction
Set the number of Pub/Sub ordering keys used to spread ordered work while retaining per-key ordering. orderingKeyShards below 1 or unordered=true requests unordered replacement, which keep_pubsub_ordering rejects.
View tool JSON
{
"name": "set_ordering_key_parallelism",
"kind": "act",
"impl": "imperative",
"description": "Set the number of Pub/Sub ordering keys used to spread ordered work while retaining per-key ordering. orderingKeyShards below 1 or unordered=true requests unordered replacement, which keep_pubsub_ordering rejects.",
"inputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string",
"enum": [
"web_client",
"api_gateway",
"cloud_run_order",
"pubsub_ordered",
"cloud_run_payment",
"cloud_sql",
"memorystore",
"global_alb",
"cloud_run_europe",
"cloud_run_us",
"app_europe",
"app_us",
"cloud_sql_primary",
"cloud_sql_replica",
"clients",
"cloud_run_router",
"vertex_stable",
"vertex_rc",
"pubsub_overflow"
]
},
"orderingKeyShards": {
"type": "number",
"minimum": 0,
"maximum": 15
},
"unordered": {
"type": "boolean"
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"id",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}set_batchingAction
Configure batching and wait time for Pub/Sub or Vertex AI within the declared model limits.
View tool JSON
{
"name": "set_batching",
"kind": "act",
"impl": "imperative",
"description": "Configure batching and wait time for Pub/Sub or Vertex AI within the declared model limits.",
"inputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string",
"enum": [
"web_client",
"api_gateway",
"cloud_run_order",
"pubsub_ordered",
"cloud_run_payment",
"cloud_sql",
"memorystore",
"global_alb",
"cloud_run_europe",
"cloud_run_us",
"app_europe",
"app_us",
"cloud_sql_primary",
"cloud_sql_replica",
"clients",
"cloud_run_router",
"vertex_stable",
"vertex_rc",
"pubsub_overflow"
]
},
"maxBatch": {
"type": "number"
},
"waitMs": {
"type": "number"
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"id",
"maxBatch",
"waitMs",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}add_read_replicaAction
Add a same-region Cloud SQL read replica for the zonal failure path. Legal even when no_second_region is pinned.
View tool JSON
{
"name": "add_read_replica",
"kind": "act",
"impl": "imperative",
"description": "Add a same-region Cloud SQL read replica for the zonal failure path. Legal even when no_second_region is pinned.",
"inputSchema": {
"type": "object",
"properties": {
"id": {
"type": "string",
"enum": [
"web_client",
"api_gateway",
"cloud_run_order",
"pubsub_ordered",
"cloud_run_payment",
"cloud_sql",
"memorystore",
"global_alb",
"cloud_run_europe",
"cloud_run_us",
"app_europe",
"app_us",
"cloud_sql_primary",
"cloud_sql_replica",
"clients",
"cloud_run_router",
"vertex_stable",
"vertex_rc",
"pubsub_overflow"
]
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"id",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}fail_regionAction
Fail one configured region while keeping its reference topology visible.
View tool JSON
{
"name": "fail_region",
"kind": "act",
"impl": "imperative",
"description": "Fail one configured region while keeping its reference topology visible.",
"inputSchema": {
"type": "object",
"properties": {
"region": {
"type": "string",
"enum": [
"europe-west2",
"us-east4"
]
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"region",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}restore_regionAction
Restore one failed region and its placed services.
View tool JSON
{
"name": "restore_region",
"kind": "act",
"impl": "imperative",
"description": "Restore one failed region and its placed services.",
"inputSchema": {
"type": "object",
"properties": {
"region": {
"type": "string",
"enum": [
"europe-west2",
"us-east4"
]
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"region",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}set_region_traffic_splitAction
Set the primary-region traffic allocation.
View tool JSON
{
"name": "set_region_traffic_split",
"kind": "act",
"impl": "imperative",
"description": "Set the primary-region traffic allocation.",
"inputSchema": {
"type": "object",
"properties": {
"primaryPercent": {
"type": "number"
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"primaryPercent",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}set_latency_based_routingAction
Enable or disable latency-based routing. Traffic shifts toward the faster healthy region or sheds a slow release candidate.
View tool JSON
{
"name": "set_latency_based_routing",
"kind": "act",
"impl": "imperative",
"description": "Enable or disable latency-based routing. Traffic shifts toward the faster healthy region or sheds a slow release candidate.",
"inputSchema": {
"type": "object",
"properties": {
"enabled": {
"type": "boolean"
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"enabled",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}set_model_traffic_splitAction
Set the release-candidate share of model traffic. The healthy default is a 20% canary.
View tool JSON
{
"name": "set_model_traffic_split",
"kind": "act",
"impl": "imperative",
"description": "Set the release-candidate share of model traffic. The healthy default is a 20% canary.",
"inputSchema": {
"type": "object",
"properties": {
"newModelPercent": {
"type": "number"
},
"expectedVersion": {
"type": "number"
}
},
"required": [
"newModelPercent",
"expectedVersion"
],
"additionalProperties": false
},
"page": "/"
}