
Mcp Vmanomaly
Connects Claude to VictoriaMetrics' vmanomaly anomaly detection service through its REST API. You get tools to check server health, list and configure detection models like zscore_online and prophet, generate complete YAML configs, and create vmalert rules based on anomaly scores. Ships with embedded searchable documentation so it works offline. Useful when you're tuning anomaly detection pipelines and want to iterate on model configs or alert thresholds conversationally instead of editing YAML by hand. Requires vmanomaly 1.28.3 or later running somewhere accessible. Available as a Go binary, Docker image, or source build with stdio and HTTP transport options.
- GitHub stars
- 11
- tools
- 19
- auth
- Secret needed
- transport
- STDIO
- env vars
- 12
- last updated
- Nov 4, 2026
Tools
19 toolsPublic tool metadata for what this MCP can expose to an agent.
vmanomaly_health_checkCheck vmanomaly server health status
Check vmanomaly server health status
No parameter schema in public metadata yet.
vmanomaly_get_buildinfoGet build information (version, build time, Go version)
Get build information (version, build time, Go version)
No parameter schema in public metadata yet.
vmanomaly_get_server_queriesGet configured server query aliases and expressions
Get configured server query aliases and expressions
No parameter schema in public metadata yet.
vmanomaly_get_metricsGet vmanomaly server metrics in Prometheus format
Get vmanomaly server metrics in Prometheus format
No parameter schema in public metadata yet.
vmanomaly_estimate_deployment_resourcesDeployment CPU/RAM/disk and composable stage estimates
Deployment CPU/RAM/disk and composable stage estimates
No parameter schema in public metadata yet.
vmanomaly_estimate_inference_capacityApproximate active models per inference interval at fixed CPUs, optionally RAM constrained
Approximate active models per inference interval at fixed CPUs, optionally RAM constrained
No parameter schema in public metadata yet.
vmanomaly_get_deployment_sizing_profilesDiscover supported sizing profiles and their model parameters
Discover supported sizing profiles and their model parameters
No parameter schema in public metadata yet.
vmanomaly_list_modelsList models exposed to VMUI and other UI-oriented flows
List models exposed to VMUI and other UI-oriented flows
No parameter schema in public metadata yet.
vmanomaly_get_server_modelsGet configured server models and their query attachments
Get configured server models and their query attachments
No parameter schema in public metadata yet.
vmanomaly_get_model_schemaGet JSON schema for a specific model type
Get JSON schema for a specific model type
No parameter schema in public metadata yet.
vmanomaly_validate_model_configValidate model configuration before using it
Validate model configuration before using it
No parameter schema in public metadata yet.
vmanomaly_validate_configValidate complete vmanomaly YAML configuration
Validate complete vmanomaly YAML configuration
No parameter schema in public metadata yet.
vmanomaly_search_docsFull-text search across vmanomaly documentation with fuzzy matching
Full-text search across vmanomaly documentation with fuzzy matching
No parameter schema in public metadata yet.
vmanomaly_check_compatibilityCheck if persisted state is compatible with runtime version
Check if persisted state is compatible with runtime version
No parameter schema in public metadata yet.
vmanomaly_generate_alert_ruleGenerate VMAlert rule YAML for anomaly score alerting
Generate VMAlert rule YAML for anomaly score alerting
No parameter schema in public metadata yet.
vmanomaly_timeseries_characteristicsProfile sampled query results for trends, seasonalities, spikiness, and gaps
Profile sampled query results for trends, seasonalities, spikiness, and gaps
No parameter schema in public metadata yet.
vmanomaly_create_autotune_taskStart tuning one requested model class on sampled series
Start tuning one requested model class on sampled series
No parameter schema in public metadata yet.
vmanomaly_get_autotune_taskPoll autotune progress and retrieve a completed recommendation
Poll autotune progress and retrieve a completed recommendation
No parameter schema in public metadata yet.
vmanomaly_cancel_autotune_taskRequest cooperative cancellation of an autotune task
Request cooperative cancellation of an autotune task
No parameter schema in public metadata yet.
Sponsors
Configuration
VMANOMALY_ENDPOINT*vmanomaly server endpoint URL (e.g., http://localhost:8490)
VMANOMALY_BEARER_TOKENsecretBearer token for authenticating with vmanomaly API
VMANOMALY_BEARER_TOKEN_FILEPath to a file containing a bearer token; mutually exclusive with VMANOMALY_BEARER_TOKEN
VMANOMALY_HEADERSCustom HTTP headers for vmanomaly requests (comma-separated key=value pairs, e.g., X-Custom-Header=value1,X-Another=value2)
MCP_SERVER_MODEMCP server mode: stdio (default), sse, or http
MCP_LISTEN_ADDRListen address for SSE/HTTP modes (default: localhost:8080)
MCP_LOG_LEVELLog level: debug, info (default), warn, or error
MCP_LOG_FILELog file path (empty = stderr)
MCP_ENABLED_TOOLSComma-separated positive allowlist of tools; empty enables all registered tools
MCP_DISABLED_TOOLSComma-separated denylist of tools; takes precedence over MCP_ENABLED_TOOLS
MCP_HEARTBEAT_INTERVALHeartbeat interval for HTTP mode (default: 30s)
MCP_DISABLE_RESOURCESDisable embedded documentation resources; documentation search remains available
Readme
MCP Server for vmanomaly
The implementation of Model Context Protocol (MCP) server for vmanomaly - VictoriaMetrics Anomaly Detection product.
This provides seamless integration with vmanomaly REST API and documentation for AI-assisted anomaly detection, model management, and observability insights.
Features
Details
- Repository
- github.com/victoriametrics/mcp-vmanomaly
- Package
ghcr.io/victoriametrics/mcp-vmanomaly:0.6.0- Transport
- STDIO
- Auth
- Required
- Registry
- MCP registry, active