Skip to main content

Resources

The Prometheus MCP Server provides 14 MCP Resources for real-time data streams. Resources are read-only — they provide context for AI agents to make informed decisions before using state-changing tools.

Available Resources​

Resource URIDescription
prom://system/backendsAll known backends with health status — start here
prom://system/backends/{backend_id}Detailed capabilities, runtime info, and health for one backend
prom://config/runtimeSanitized runtime config: global settings, remote-write targets, TSDB stats
prom://topology/servicesService catalog derived from scrape targets with health status
prom://topology/services/{job}/metricsAll metrics emitted by a specific service/job
prom://topology/failed_targetsAggregated view of failed/down scrape targets
prom://metadata/catalogMetric names with type and HELP text — prevents hallucination
prom://schema/label_valuesPer-metric label values snapshot
prom://tsdb/cardinalityTSDB cardinality overview and top-N high-cardinality metrics
prom://rules/groupsAlerting and recording rule group inventory
prom://kubernetes/prometheusrulesPrometheusRule CRDs with K8s metadata (name, namespace, labels)
prom://exporters/catalogBuilt-in exporter catalog with types, ports, images
prom://best-practicesPrometheus best practices reference
prom://onboarding-guideStep-by-step onboarding guide

prom://system/backends​

Cluster-wide listing of all configured Prometheus backends with health status. Always read this first to discover available backend_id values.

Use Case:

"What Prometheus backends are available?"
"Check if our production Prometheus is healthy."

prom://system/backends/{backend_id}​

Detailed information for a specific backend — build info, feature flags, storage retention, and runtime configuration.

URI Example: prom://system/backends/prod

Use Case:

"Show me the details of our production Prometheus backend."
"What features are enabled on our Thanos instance?"

prom://topology/services​

Service catalog derived from active scrape targets. Shows which jobs exist, their health status (up/down), and target count.

Use Case:

"What services are being monitored by Prometheus?"
"List all scrape jobs and their health status."

prom://topology/services/{job}/metrics​

All metrics emitted by a specific job, including metric type (counter, gauge, histogram, summary) and HELP text.

URI Example: prom://topology/services/api-server/metrics

Use Case:

"What metrics does the api-server expose?"
"Show me all histogram metrics from the frontend service."

prom://topology/failed_targets​

Aggregated view of all failed or down scrape targets — essential for troubleshooting.

Use Case:

"Which scrape targets are currently failing?"
"Show me all down targets — we might have a networking issue."

prom://metadata/catalog​

Complete metric name catalog with types and HELP text. Prevents the AI from hallucinating metric names.

Use Case:

"What metrics are available that relate to HTTP requests?"
"Search for any metrics containing 'memory' in their name."

prom://tsdb/cardinality​

TSDB cardinality overview: total active series, top-N high-cardinality metrics, and label pair counts.

Use Case:

"How many active series does our Prometheus have?"
"Which metrics are consuming the most TSDB storage?"

prom://rules/groups​

Inventory of all loaded alerting and recording rule groups across backends.

Use Case:

"List all alerting rule groups in our production Prometheus."
"How many rules do we have in total?"

prom://kubernetes/prometheusrules​

All PrometheusRule CRDs in the cluster with Kubernetes metadata — name, namespace, labels, and group summaries. Required for safe prom_upsert_rule_group operations in k8s_crd mode.

Use Case:

"Find the CRD that owns the alertmanager.rules group."
"List all PrometheusRule resources in the monitoring namespace."
warning

Always cross-reference this resource with prom://rules/groups before calling prom_upsert_rule_group. An incorrect namespace will silently create a duplicate CRD instead of patching the existing one.

prom://exporters/catalog​

Built-in catalog of 19 supported exporters with types, default ports, Docker images, and supported environments.

Use Case:

"What exporters are available for PostgreSQL?"
"Show me all exporters that support DaemonSet deployment."

prom://best-practices / prom://onboarding-guide​

Static documentation resources providing Prometheus best practices and step-by-step onboarding guidance.

Use Case:

"What are the best practices for Prometheus labeling?"
"Show me the onboarding guide for new applications."

Using Resources​

Resources are accessed via the MCP protocol using their URI. AI agents read resources for context before making tool calls.

Client: What Prometheus backends are available?
Agent: [Reads prom://system/backends]
       [Returns list of backends with health status]
Client: Why is my api-server job showing as down?
Agent: [Reads prom://topology/failed_targets]
       [Returns failed targets with error reasons]

Next Steps​