Prometheus Api

What it does

This skill enables AI agents to interact with Prometheus monitoring systems via their HTTP API at /api/v1. It provides access to various endpoints for querying metrics, formatting PromQL queries, discovering series and labels, and retrieving metadata and status information about the Prometheus server itself. The responses are formatted as JSON, including potential error details.

When to use it

• Troubleshooting performance issues by querying specific metrics in real-time.
• Analyzing trends over time using range queries to identify anomalies.
• Discovering available labels within a Prometheus environment.
• Retrieving the status of targets and alerts managed by Prometheus.
• Formatting complex PromQL queries for clarity or optimization.

Key capabilities

• Instant Query: Retrieves metric values at a specific point in time using curl.
• Range Query: Fetches metric data over a defined time range with specified step intervals.
• PromQL Formatting: Formats PromQL queries to improve readability or identify errors.
• Series Discovery: Finds series of metrics based on label matching.
• Label Listing & Value Retrieval: Lists available labels and retrieves their values.
• Metadata Access: Retrieves information about targets, alerts, rules, and the Prometheus server's configuration and runtime status.

Example prompts

• "Run an instant query for CPU usage on my Prometheus server."
• "Give me a range query of memory utilization over the last 5 minutes."
• "Format this PromQL query: up{job="my-app"} == 1"
• "List all available labels related to network traffic."

Tips & gotchas

• Requires access to a Prometheus server running on port 9090 (or the configured port).
• The skill returns raw HTTP responses in JSON format. You may need additional processing to extract and interpret the data.
• Admin endpoints require --web.enable-admin-api to be enabled on the Prometheus server.