Available Tools
Complete reference for all MCP tools provided by the Prometheus MCP Server. Tools are organized by category — 28 tools across 6 categories.
PromQL Query Engine
Tools for safely querying Prometheus with counter enforcement and automatic downsampling.
| Tool | Description |
|---|---|
prom_validate_promql | Check PromQL syntax before executing |
prom_query_instant | Execute a point-in-time PromQL query with counter enforcement |
prom_query_range | Execute a time-range PromQL query with automatic downsampling |
prom_explore_labels | Discover label names and top values for a given metric |
prom_suggest_promql | Generate a PromQL expression from natural language intent |
prom_query_instant
Executes an instant (point-in-time) query. Enforces counter semantics — raw counter queries are blocked unless allow_raw_counters=true.
"What is the current request rate for http_requests_total?"
"Show CPU usage across all pods in the production namespace."
"Query up{job='api-server'} — is my service being scraped?"
prom_query_range
Executes a range query with automatic step calculation. Downsamples results to ~200 points per series to protect LLM context windows.
"Show me the request rate for http_requests_total over the last hour."
"Graph CPU usage for the last 24 hours."
Safety guardrails
| Guardrail | Rule | Override |
|---|---|---|
| Counter Enforcement | Counters must use rate() or increase() | allow_raw_counters=true |
| Auto-Downsampling | Range queries capped at ~200 points/series | max_points_per_series param |
| Validate First | Complex queries should be checked via prom_validate_promql | — |
Application Onboarding
Tools for instrumenting applications and wiring them into Prometheus.
| Tool | Description |
|---|---|
prom_recommend_instrumentation | Recommend direct instrumentation vs exporter vs builtin metrics |
prom_test_endpoint | Validate that an endpoint exposes valid Prometheus/OpenMetrics metrics |
prom_apply_servicemonitor | Generate and apply a ServiceMonitor CRD for Prometheus Operator |
prom_apply_probe | Generate and apply a Probe CRD for synthetic endpoint monitoring |
prom_manage_file_sd | Add or remove static targets in a file_sd_configs JSON file |
prom_recommend_instrumentation
Recommends the best monitoring strategy based on workload type, language, and environment.
"What's the best way to monitor my Python Flask app on Kubernetes?"
"Should I use an exporter or direct instrumentation for our Go service?"
Returns one of: direct_instrumentation, exporter, or builtin_metrics (e.g., Spring Boot Actuator/Micrometer).
prom_apply_servicemonitor
Creates a ServiceMonitor CRD with auto-detected Prometheus Operator selector labels. Mutates the cluster.
"Wire my api-server service in the default namespace to Prometheus."
"Create a ServiceMonitor for postgres-exporter in monitoring namespace."
prom_apply_probe
Creates a Probe CRD for synthetic endpoint monitoring using Blackbox exporter.
"Monitor https://talkops.ai availability from inside the cluster."
"Set up a probe for our health check endpoint at /healthz."
prom_manage_file_sd
Manages static targets for VM/legacy environments. Supports add and remove sub-actions. Optionally triggers POST /-/reload.
"Add my VM host 10.0.1.5:9100 to Prometheus file_sd targets."
"Remove the old database target from file_sd."
Exporter Lifecycle
Tools for deploying and managing Prometheus exporters on Kubernetes.
| Tool | Description |
|---|---|
prom_recommend_exporter | Get exporter recommendations for a specific service type |
prom_install_exporter | Deploy an exporter to Kubernetes (Deployment/DaemonSet + Service + RBAC) |
prom_uninstall_exporter | Remove an exporter from Kubernetes |
prom_verify_exporter | End-to-end health check: scrape endpoint + check Prometheus up{} series |
prom_install_exporter
Deploys a complete exporter stack to Kubernetes — Deployment (or DaemonSet), Service, and optionally RBAC roles, ConfigMaps, and additional manifests. Mutates the cluster.
"Deploy a postgres_exporter in the monitoring namespace for our PostgreSQL database."
"Install node_exporter as a DaemonSet across all nodes."
"Set up a blackbox_exporter for synthetic endpoint monitoring."
Supported exporters (19 total)
| Exporter | Default Port | Scope | K8s Nuances |
|---|---|---|---|
node_exporter | 9100 | DaemonSet | — |
kube-state-metrics | 8080 | Deployment | Requires RBAC |
postgres_exporter | 9187 | Sidecar | Sidecar support |
mysqld_exporter | 9104 | Sidecar | Sidecar support |
mongodb_exporter | 9216 | Deployment | Sidecar support |
redis_exporter | 9121 | Sidecar | Sidecar support |
nginx_exporter | 9113 | Sidecar | Sidecar support |
kafka_exporter | 9308 | Deployment | — |
rabbitmq_exporter | 9419 | Deployment | — |
elasticsearch_exporter | 9114 | Deployment | — |
blackbox_exporter | 9115 | Deployment | Requires ConfigMap |
snmp_exporter | 9116 | Deployment | Requires ConfigMap |
statsd_exporter | 9102 | Deployment | UDP Service + ConfigMap |
jmx_exporter | 5556 | Sidecar | ConfigMap + Sidecar |
aws_cloudwatch_exporter | 5000 | Deployment | ConfigMap |
apache_exporter | 9117 | Sidecar | Sidecar support |
php_fpm_exporter | 9253 | Sidecar | Sidecar support |
memcached_exporter | 9150 | Deployment | — |
windows_exporter | 9182 | DaemonSet | VM only |
Rule Management & Authoring
Tools for creating, validating, and managing alerting and recording rules.
| Tool | Description |
|---|---|
prom_get_rule_group | Get a single rule group by name with full rule details |
prom_upsert_rule_group | Create or update a rule group (YAML, CRD, or HTTP Ruler) |
prom_delete_rule_group | Delete a rule group |
prom_describe_alert_rule | Human-readable explanation of an alerting rule |
prom_draft_alert_rule | Generate an alert rule from natural language intent |
prom_tune_alert_rule | Suggest rule adjustments based on firing history |
prom_draft_alert_rule
Generates a complete alerting rule definition from natural language.
"Draft an alert for when the 5xx error rate exceeds 5% for 5 minutes."
"Create a recording rule for pre-aggregated request rates by service."
prom_upsert_rule_group
Creates or updates a rule group. Supports three storage modes:
| Storage Mode | Description |
|---|---|
yaml | Writes YAML rule file to disk |
k8s_crd | Patches a PrometheusRule CRD in Kubernetes |
http_ruler | POSTs to the HTTP Ruler API (Thanos/Mimir/Cortex) |
When using k8s_crd mode, always cross-reference prom://kubernetes/prometheusrules to discover the exact CRD name and namespace. An incorrect namespace will silently create a duplicate CRD instead of patching the existing one.
Rule Simulation & Testing
Tools for validating and testing rules before deploying them.
| Tool | Description |
|---|---|
prom_check_rule_group | Validate rule group syntax via promtool |
prom_run_rule_tests | Run promtool test rules with synthetic test scenarios |
prom_simulate_firing_synthetic | Run synthetic alert firing test via promtool |
prom_simulate_firing_historical | Evaluate alert expression against real historical data |
prom_analyze_firing_history | Analyze alert firing frequency and duration for tuning |
prom_simulate_firing_historical
The most powerful testing tool — evaluates an alert expression against your real historical data to see if it would have fired over a given time period.
"If I set the error rate threshold to 3%, would this alert have fired in the last 7 days?"
"Simulate the high memory alert against last week's production data."
TSDB FinOps & Optimization
Tools for analyzing and optimizing Prometheus storage costs and cardinality.
| Tool | Description |
|---|---|
prom_optimize_cardinality | Analyze top-N cardinality metrics and recommend optimization strategies |
prom_plan_relabel | Generate metric_relabel_configs YAML to drop/keep labels |
prom_create_recording_rule | Generate recording rule group YAML |
prom_configure_remote_write | Generate remote_write config YAML for long-term storage |
All FinOps tools generate configuration YAML — they do not apply changes to a running Prometheus instance. Users must manually add the YAML to their Prometheus configuration files.
"Analyze cardinality hotspots and recommend optimization strategies."
"Generate a relabel config to drop the pod_id label from http_requests_total."
"Configure remote-write to our Thanos receive endpoint."
Tool Availability Summary
| Category | Total Tools |
|---|---|
| PromQL Query Engine | 5 |
| Application Onboarding | 5 |
| Exporter Lifecycle | 4 |
| Rule Management & Authoring | 6 |
| Rule Simulation & Testing | 5 |
| TSDB FinOps & Optimization | 4 |
| Total | 29 |
Next Steps
- Resources — 14 MCP resources for discovery and monitoring
- Common Workflows — Step-by-step workflow guides
- Examples — Quick reference and natural language prompts