Cloud World Model
Simulate cloud infrastructure and estimate cost before deploying or changing infrastructure.
Should I use this
Quality & Safety
Findings (10)
- LOWin simulation.from_input
- LOWin simulation.inject_failure
- LOWin simulation.recover_resource
- LOWin simulation.delete
- LOWin scenario.list
- LOWin scenario.get
- LOWin simulation.create
- LOWin simulation.step
- LOWin simulation.metrics
- LOWin simulation.inject_traffic
Based on automated analysis of tool definitions and protocol compliance.
Context Cost
This is the approximate number of tokens consumed each time the server's tools are loaded into a model's context. Higher counts reduce the attention available for other tasks.
Install
One-Click Install
Add this to your `claude_desktop_config.json` file:
{
"mcpServers": {
"cloud-world-model": {
"url": "https://www.cloudworldmodel.ai/mcp"
}
}
}Remote endpoints
https://www.cloudworldmodel.ai/mcpstreamable-httpWhat it can do
Tool inventory
Tools (10)
🟢simulation.from_input(format, content, filename, files, environment, ...)
Key/content/target;cloud terraform-plan:no host;DaemonSet:placement.workers,placement.podSlotsPerWorker;GPU;simulation.step Read resource cwm://guides/public/simulation.from_input
Input Schema
{
"type": "object",
"properties": {
"format": {
"enum": [
"compose",
"kubernetes",
"terraform-plan"
]
},
"content": {
"type": "string",
"minLength": 1,
"maxLength": 262144
},
"filename": {
"type": "string",
"minLength": 1,
"maxLength": 256
},
"files": {
"type": "object",
"additionalProperties": {
"type": "string",
"maxLength": 262144
},
"default": {}
},
"environment": {
"type": "object",
"additionalProperties": {
"type": "string",
"maxLength": 8192
},
"default": {}
},
"activeProfiles": {
"type": "array",
"items": {
"type": "string",
"pattern": "^[a-zA-Z0-9][a-zA-Z0-9_.-]{0,63}$"
},
"maxItems": 32,
"default": []
},
"diagnosticLoads": {
"type": "array",
"items": {
"type": "object",
"properties": {
"service": {
"$ref": "#/$defs/a"
},
"backgroundCpuMilliCoresPerReplica": {
"type": "number",
"minimum": 0,
"maximum": 1000000
}
},
"required": [
"service",
"backgroundCpuMilliCoresPerReplica"
],
"additionalProperties": false
},
"maxItems": 256
},
"host": {
"type": "object",
"properties": {
"provider": {
"enum": [
"digitalocean",
"aws",
"azure",
"gcp",
"oci"
]
},
"size": {
"$ref": "#/$defs/b"
},
"region": {
"$ref": "#/$defs/b"
},
"reserveGiB": {
"type": "number",
"minimum": 0,
"maximum": 4096,
"default": 0.5
},
"swapGiB": {
"$ref": "#/$defs/c"
},
"nodeSwapGiB": {
"type": "object",
"additionalProperties": {
"$ref": "#/$defs/c"
}
},
"configurationFiles": {
"type": "array",
"items": {
"type": "string",
"minLength": 1,
"maxLength": 256
},
"maxItems": 32
}
},
"required": [
"provider",
"size",
"region"
],
"additionalProperties": false
},
"gpu": {
"type": "object"
},
"target": {
"type": "object",
"properties": {
"service": {
"$ref": "#/$defs/a"
},
"kind": {
"enum": [
"entry",
"background"
]
},
"resourceKind": {
"enum": [
"workload",
"Service",
"Ingress"
]
}
},
"required": [
"service",
"kind"
],
"additionalProperties": false
},
"traffic": {
"type": "object",
"properties": {
"initialRps": {
"type": "number",
"minimum": 0,
"maximum": 500000,
"default": 100
},
"cpuMilliCoresPerRequest": {
"type": "number",
"exclusiveMinimum": 0,
"maximum": 1000000,
"default": 1
}
},
"additionalProperties": false,
"default": {}
},
"dependencies": {
"type": "object",
"properties": {
"startupAsRuntime": {
"type": "boolean",
"default": false
},
"overrides": {
"type": "array",
"items": {
"type": "object",
"properties": {
"service": {
"$ref": "#/$defs/a"
},
"dependency": {
"$ref": "#/$defs/a"
},
"criticality": {
"enum": [
"critical",
"optional",
"startup-only"
]
}
},
"required": [
"service",
"dependency",
"criticality"
],
"additionalProperties": false
},
"maxItems": 256,
"default": []
}
},
"additionalProperties": false,
"default": {}
},
"placement": {
"type": "object",
"properties": {
"workers": {
"type": "array",
"items": {
"type": "object",
"properties": {
"id": {
"$ref": "#/$defs/b"
},
"zone": {
"$ref": "#/$defs/b"
},
"podSlots": {
"type": "integer",
"minimum": 0,
"maximum": 1000
},
"node": {
"type": "object"
},
"swapGiB": {
"$ref": "#/$defs/c"
},
"instanceType": {
"$ref": "#/$defs/b"
},
"region": {
"$ref": "#/$defs/b"
}
},
"required": [
"id",
"podSlots"
],
"additionalProperties": false
},
"minItems": 1,
"maxItems": 128
},
"podSlotsPerWorker": {
"type": "integer",
"minimum": 1,
"maximum": 1000
},
"drains": {
"type": "array",
"items": {
"type": "object",
"properties": {
"workerId": {
"$ref": "#/$defs/b"
},
"step": {
"type": "integer",
"minimum": 0,
"maximum": 100000
}
},
"required": [
"workerId",
"step"
],
"additionalProperties": false
},
"maxItems": 128,
"default": []
},
"startupDelaySeconds": {
"type": "number",
"minimum": 0,
"maximum": 3600,
"default": 0
},
"terminationGracePeriodSeconds": {
"type": "number",
"minimum": 0,
"maximum": 3600,
"default": 30
}
},
"additionalProperties": false
},
"name": {
"type": "string",
"minLength": 1,
"maxLength": 128,
"default": "Imported stack"
},
"seed": {
"type": "integer",
"minimum": 0,
"maximum": 4294967295,
"default": 1
}
},
"required": [
"format",
"content",
"target"
],
"additionalProperties": false,
"$defs": {
"a": {
"type": "string",
"minLength": 1,
"maxLength": 512
},
"b": {
"type": "string",
"minLength": 1,
"maxLength": 128
},
"c": {
"type": "number",
"minimum": 0,
"maximum": 4096
}
}
}🟢simulation.inject_failure(simulationId, resourceName, resourceId, nodePoolId, nodePoolName, ...)
Requires simulationId and resourceId from simulation.create. Injects a bounded resource failure; returns updated state. Next tool is simulation.step, then simulation.recover_resource and simulation.metrics to confirm authoritative health. No API key required; demo limits apply. Read resource cwm://guides/public/simulation.inject_failure
Input Schema
{
"type": "object",
"properties": {
"simulationId": {
"type": "string",
"description": "ID returned by simulation.create."
},
"resourceName": {
"type": "string"
},
"resourceId": {
"type": "string"
},
"nodePoolId": {
"type": "string",
"minLength": 1
},
"nodePoolName": {
"type": "string",
"minLength": 1
},
"type": {
"type": "string",
"enum": [
"cdn_cache_loss",
"az_outage"
],
"description": "Only cdn_cache_loss or az_outage, with an explicit CDN-capable network resourceId or resourceName; no random/node-pool selectors. Cache loss forwards misses; az_outage blocks the edge. Omit type and duration for legacy quick failures."
},
"duration": {
"type": "integer",
"minimum": 1,
"maximum": 86400,
"description": "Only with explicit type cdn_cache_loss or az_outage and CDN-capable network resourceId or resourceName. Integer 1–86400 simulation steps, not seconds; default 30. No random/node-pool selectors; omit for legacy quick failures."
}
},
"additionalProperties": false
}🟢simulation.recover_resource(simulationId, resourceName, resourceId)
Requires simulationId and resourceId from simulation.create. Requests reversible resource recovery; success does not prove healthy. Next tool is simulation.step, then simulation.metrics to inspect resource health. No API key required; demo limits apply. Read resource cwm://guides/public/simulation.recover_resource
Input Schema
{
"type": "object",
"properties": {
"simulationId": {
"type": "string",
"description": "ID returned by simulation.create."
},
"resourceName": {
"type": "string"
},
"resourceId": {
"type": "string"
}
},
"additionalProperties": false
}🔴simulation.delete(simulationId)
The likely next tool is simulation.create to use the freed slot. Prerequisite: a simulationId from simulation.create, or an active simulation in the preserved MCP session. No API key required; demo limits apply. Read resource cwm://guides/public/simulation.delete
Input Schema
{
"type": "object",
"properties": {
"simulationId": {
"type": "string",
"description": "ID returned by simulation.create."
}
},
"additionalProperties": false
}🟢scenario.list(limit, cursor, detailed, provider, category, ...)
No prerequisites. Returns paginated scenarios cards with scenarioId (alias of id), name, summary and available provider/category/difficulty. Pass scenarioId to scenario.get. Filters precede stable ID paging: limit default 8, maximum 20; pass nextCursor as cursor until null. detailed=true restores rich cards. Next tool is scenario.get, then simulation.create with scenarioId. No API key required; demo limits apply. Read resource cwm://guides/public/scenario.list
Input Schema
{
"type": "object",
"properties": {
"limit": {
"type": "integer",
"default": 8,
"minimum": 1,
"maximum": 20
},
"cursor": {
"type": "string",
"maxLength": 256
},
"detailed": {
"type": "boolean",
"default": false
},
"provider": {
"type": "string",
"enum": [
"aws",
"gcp",
"azure",
"oci",
"digitalocean"
]
},
"category": {
"type": "string",
"minLength": 1
},
"difficulty": {
"type": "string",
"enum": [
"beginner",
"intermediate",
"advanced"
]
}
},
"additionalProperties": false
}Output Schema
{
"type": "object",
"properties": {
"scenarios": {
"type": "array",
"items": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"scenarioId": {
"type": "string"
},
"name": {
"type": "string"
},
"provider": {
"type": "string"
},
"category": {
"type": "string"
},
"difficulty": {
"type": "string"
}
},
"required": [
"id",
"scenarioId",
"name",
"category"
],
"additionalProperties": true
}
},
"total": {
"type": "integer",
"minimum": 0
},
"limit": {
"type": "integer",
"minimum": 1,
"maximum": 20
},
"nextCursor": {
"type": [
"string",
"null"
]
}
},
"required": [
"scenarios",
"total",
"limit",
"nextCursor"
],
"additionalProperties": true,
"$schema": "http://json-schema.org/draft-07/schema#"
}🟢scenario.get(scenarioId)
Requires scenarioId returned by scenario.list (also available as id). Returns the complete selected graph, presets, assumptions and phase details. Next tool is simulation.create with scenarioId or the hydrated resources/connections. No API key required; demo limits apply. Read resource cwm://guides/public/scenario.get
Input Schema
{
"type": "object",
"properties": {
"scenarioId": {
"type": "string",
"minLength": 1,
"description": "Scenario ID from scenario.list."
}
},
"required": [
"scenarioId"
],
"additionalProperties": false
}🟡simulation.create(costAssumptions, name, appWeight, description, scenarioId, ...)
Create from scenarioId (scenario.list → scenario.get), or resources/connections for a custom graph. Returns id for simulation.step. Never blindly retry a timeout: creation may have succeeded. No API key required; demo limits apply. Read resource cwm://guides/public/simulation.create GCP gcpFitPrediction: holdout-validated for one us-central1 topology at 300–500 RPS; see guide.
Input Schema
{
"type": "object",
"properties": {
"costAssumptions": {
"type": "object",
"additionalProperties": true
},
"name": {
"type": "string",
"maxLength": 120
},
"appWeight": {
"type": "string",
"enum": [
"lean",
"typical",
"heavy"
]
},
"description": {
"type": "string",
"maxLength": 500
},
"scenarioId": {
"type": "string",
"minLength": 1,
"description": "Scenario ID from scenario.list."
},
"scenarioOverrides": {},
"resources": {
"type": "array",
"items": {
"type": "object",
"properties": {
"id": {
"type": "string"
},
"type": {
"type": "string",
"enum": [
"compute",
"database",
"storage",
"network",
"cache",
"queue",
"kubernetes"
]
},
"name": {
"type": "string"
},
"characteristics": {
"type": "object",
"properties": {
"databaseConnectionModel": {
"type": "string",
"const": "held-pools",
"description": "Physical database opt-in: held-pools uses current running host replicas for named/legacy pools; static binding preserves independent consumers. Explicit managedPoolId deduplicates backend pools; dedicated proxy server slots never multiply by tasks. Inspect runtimeConnections, pools and connectionLimitProvenance."
},
"maxConnections": {
"type": "integer",
"exclusiveMinimum": 0,
"maximum": 1000000,
"description": "Fixed concurrent DB connection budget; Aurora Serverless defaults from maximum configured ACU."
},
"appDbPoolSize": {
"type": "integer",
"exclusiveMinimum": 0,
"description": "Owned typical: pool size 250 on each app."
},
"databaseConsumers": {
"type": "array",
"description": "Named logical pools, not billed resources. On the target database select held-pools to assume full pools retained by current running host replicas. Static binding preserves independent replicas. Explicit managedPoolId/mode shares backend slots once; proxy is dedicated. Full constraints and transaction client/server limits: linked tool guide. Results: metrics.databases[].runtimeConnections, pools, firstSaturation and connectionLimitProvenance; ASSUMED/MODELED, not measured sessions.",
"items": {
"type": "object",
"additionalProperties": true,
"properties": {
"id": {
"type": "string",
"pattern": "^[a-zA-Z0-9][a-zA-Z0-9_.:/#-]{0,511}$"
},
"name": {
"type": "string",
"minLength": 1,
"maxLength": 128
},
"databaseId": {
"$ref": "#/properties/resources/items/properties/characteristics/properties/databaseConsumers/items/properties/id"
},
"poolSize": {
"type": "integer",
"exclusiveMinimum": 0,
"maximum": 1000000
},
"replicas": {
"type": "integer",
"exclusiveMinimum": 0,
"maximum": 1000,
"default": 1
},
"replicaBinding": {
"type": "string",
"enum": [
"host-runtime",
"static"
]
},
"managedPoolId": {
"$ref": "#/properties/resources/items/properties/characteristics/properties/databaseConsumers/items/properties/id"
}
}
}
},
"gpuReadiness": {
"type": "object",
"description": "EKS assumptions, not measured performance. initialReadyNodes required; warmSpareNodes adds ready reserve. autoscalerReaction/nodeProvision/imagePull/modelLoad: {seconds,provenance}. capacity: {unit:'requests-per-second'|'concurrent-calls',perNode,provenance}; concurrent-calls needs meanCallDurationSeconds. servedRequestP99Ms: {value,provenance}. provenance: user-supplied|estimated. Missing phases/capacity stay not_evaluated; omit for legacy behavior. Full constraints: linked guide."
},
"nodePools": {
"type": "array",
"items": {
"type": "object",
"additionalProperties": true
},
"description": "EKS GPU pools: id/name, size, nodeCount/minNodes/maxNodes; each has gpuReadiness as above."
}
},
"additionalProperties": true
}
},
"required": [
"id",
"type",
"name"
],
"additionalProperties": true
}
},
"connections": {
"type": "array",
"items": {
"type": "object",
"properties": {
"sourceId": {
"type": "string"
},
"targetId": {
"type": "string"
}
},
"required": [
"sourceId",
"targetId"
],
"additionalProperties": true
}
},
"traffic": {
"type": "number"
},
"seed": {
"type": "integer",
"minimum": 0
},
"autoscalingConfig": {
"type": "object",
"properties": {
"scaleOutCpuThreshold": {
"type": "number",
"minimum": 0
},
"scaleInCpuThreshold": {
"type": "number",
"minimum": 0
},
"scaleOutThroughputThreshold": {
"type": "number",
"minimum": 0
},
"scaleInThroughputThreshold": {
"type": "number",
"minimum": 0
},
"scaleOutLatencyThreshold": {
"type": "number",
"minimum": 0
},
"cooldownSeconds": {
"type": "number",
"minimum": 0
},
"ecsCpuTargetTracking": {
"type": "boolean"
},
"scaleOutCooldownSeconds": {
"type": "integer",
"minimum": 0,
"maximum": 86400
},
"scaleInCooldownSeconds": {
"type": "integer",
"minimum": 0,
"maximum": 86400
},
"simulationSecondsPerStep": {
"type": "number",
"exclusiveMinimum": 0,
"maximum": 60
},
"minInstances": {
"type": "integer",
"minimum": 0
},
"maxInstances": {
"type": "integer",
"exclusiveMinimum": 0
}
},
"required": [
"scaleOutCpuThreshold",
"scaleInCpuThreshold",
"scaleOutThroughputThreshold",
"scaleInThroughputThreshold",
"scaleOutLatencyThreshold",
"cooldownSeconds",
"minInstances",
"maxInstances"
],
"additionalProperties": false,
"description": "Complete autoscaling policy, including min/max bounds, thresholds and cooldowns. Preserved without duplicate top-level bounds; explicit top-level bounds and CPU targets take precedence. Unknown fields are rejected. Pass scenario.get's config when copying its graph."
},
"maxInstances": {
"type": "integer",
"minimum": 1
},
"minInstances": {
"type": "integer",
"minimum": 1
},
"autoscalingTargetCpu": {
"type": "number",
"minimum": 0,
"maximum": 100
},
"scaleOutCpuThreshold": {
"type": "number",
"minimum": 0,
"maximum": 100
},
"scaleOutCpuPercent": {
"type": "number",
"minimum": 0,
"maximum": 100
},
"autoscaleTargetCpuPercent": {
"type": "number",
"minimum": 0,
"maximum": 100
},
"ecsCpuTargetTracking": {
"type": "boolean",
"description": "Opt into CPU-only ECS Fargate target tracking; requires an AWS Fargate compute resource. Required for scaleOutCooldownSeconds, scaleInCooldownSeconds and top-level simulationSecondsPerStep; other providers keep their existing policy."
},
"scaleOutCooldownSeconds": {
"type": "integer",
"minimum": 0,
"maximum": 86400,
"description": "ECS scale-out cooldown in simulated seconds; requires ecsCpuTargetTracking: true with an AWS Fargate compute resource. Defaults to cooldownSeconds."
},
"scaleInCooldownSeconds": {
"type": "integer",
"minimum": 0,
"maximum": 86400,
"description": "ECS scale-in cooldown in simulated seconds; requires ecsCpuTargetTracking: true with an AWS Fargate compute resource. Defaults to cooldownSeconds."
},
"simulationSecondsPerStep": {
"type": "number",
"maximum": 60,
"description": "Requires ecsCpuTargetTracking: true with an AWS Fargate compute resource. Converts ECS cooldown seconds into step counts only (default 1); not a global simulation clock."
},
"resilienceConfig": {
"type": "object",
"additionalProperties": true
},
"responseMode": {
"type": "string",
"default": "compact",
"enum": [
"compact",
"full"
]
}
},
"additionalProperties": false
}🟢simulation.step(operationId, expectedCheckpoint, simulationId, steps, traffic, ...)
Requires simulationId from simulation.create (optional in demo). Returns completedSteps/currentStep; next tool is simulation.metrics. Retry same operationId/inputs. CWM_RAMP_GAP_NOT_ERRORS="true": owned AWS lean/2-app typical-v1 (500-RPS diagnostic) returns rampAveragingShortfallPct=100*(1-throughput/offeredRps), not errors; goodputSemantics=whole_run_average_including_warmup_ramp. No API key required; demo limits apply. Read resource cwm://guides/public/simulation.step calibrationEligibility: recorded AWS selection, reasons and warnings; not measurement validation. Owned AWS calibrated outputs may include rampAveragingShortfallPct: the warmup-ramp goodput gap, reported separately and not counted as errors.
Input Schema
{
"type": "object",
"properties": {
"operationId": {
"type": "string",
"minLength": 1,
"maxLength": 128,
"description": "Optional retry ID: omit for a new ID; reuse with identical inputs replays the saved result, while changed inputs, expired results or another run return 409."
},
"expectedCheckpoint": {
"type": "object",
"additionalProperties": true,
"description": "Optional guard for new steps: omit to skip; a mismatched generation, currentStep or supplied metricId rejects the step with 409 without advancing."
},
"simulationId": {
"type": "string",
"description": "ID returned by simulation.create."
},
"steps": {
"type": "integer",
"minimum": 1,
"maximum": 500
},
"traffic": {
"type": "number",
"minimum": 0
},
"responseMode": {
"type": "string",
"default": "compact",
"enum": [
"compact",
"full"
]
}
},
"additionalProperties": false
}🟢simulation.metrics(simulationId, responseMode)
Requires simulationId from simulation.create (demo optional). Returns latest/history; full=evidence. CWM_RAMP_GAP_NOT_ERRORS="true": owned AWS lean/2-app typical-v1 (500-RPS diagnostic) returns rampAveragingShortfallPct=100*(1-throughput/offeredRps), not errors; goodputSemantics=whole_run_average_including_warmup_ramp. No API key required; demo limits apply. cwm://guides/public/simulation.metrics calibrationEligibility: recorded AWS selection, reasons and warnings; not measurement validation. GCP gcpFitPrediction: holdout-validated for one us-central1 topology at 300–500 RPS; see guide. Owned AWS calibrated outputs may include rampAveragingShortfallPct: the warmup-ramp goodput gap, reported separately and not counted as errors.
Input Schema
{
"type": "object",
"properties": {
"simulationId": {
"type": "string",
"description": "ID returned by simulation.create."
},
"responseMode": {
"type": "string",
"default": "compact",
"enum": [
"compact",
"full"
]
}
},
"additionalProperties": false
}🟢simulation.inject_traffic(simulationId, targetRps, deltaPercent, ramp, random)
Requires simulationId from simulation.create and target traffic RPS. Returns applied traffic settings. Next tool is simulation.step, then simulation.metrics to read observed latency; do not assume its direction. No API key required; demo limits apply. Read resource cwm://guides/public/simulation.inject_traffic
Input Schema
{
"type": "object",
"properties": {
"simulationId": {
"type": "string",
"description": "ID returned by simulation.create."
},
"targetRps": {
"type": "integer",
"minimum": 1
},
"deltaPercent": {
"type": "number"
},
"ramp": {
"type": "object",
"additionalProperties": true
},
"random": {
"type": "boolean",
"const": true
}
},
"additionalProperties": false
}Community
Evidence