Do I need to install any special tools?

The skill utilizes standard Bash and curl commands to interact with the Prometheus API, making it highly portable across development environments.

Does this skill work with different OpenTelemetry SDKs?

Yes, it includes a discovery script to identify active metric names regardless of whether your SDK uses older or newer semantic conventions.

What kind of metrics can I track for GenAI applications?

You can monitor token usage histograms and operation duration grouped by specific AI models and operation names like 'chat' or 'embeddings'.

Can I use this with Amazon Managed Service for Prometheus?

Yes, the skill provides specific configurations for AWS AMP, including SigV4 authentication and workspace endpoint formatting.

Observability Metrics & PromQL

Name: Observability Metrics & PromQL
Author: opensearch-project

byopensearch-project

•

분석 및 모니터링

Queries and analyzes Prometheus metrics using PromQL to monitor HTTP performance, database latency, and GenAI token usage.

This skill empowers developers to monitor and troubleshoot applications by providing ready-to-use PromQL templates for the Prometheus HTTP API. It handles metric discovery across different OpenTelemetry SDK versions and enables deep insights into request rates, error codes, p95/p99 latency percentiles, and specific GenAI model performance. Whether working with a local stack or Amazon Managed Service for Prometheus (AMP), it streamlines the process of extracting actionable observability data directly within the Claude Code environment, helping teams maintain high system availability and performance.

주요 기능

01Specialized monitoring for GenAI token usage and model operation duration

02Pre-configured templates for HTTP request rates and 5xx error rate ratios

03High-precision latency analysis for p95 and p99 percentiles by service

0421 GitHub stars

05Automated discovery of active metric names across varying OTel SDK versions

06Full support for local Prometheus and AWS Managed Service for Prometheus (AMP)

사용 사례

01Auditing database operation performance to identify slow queries and backend bottlenecks

02Diagnosing sudden spikes in HTTP error rates or latency across distributed microservices

03Monitoring GenAI costs and efficiency by tracking token usage per model and operation

주요 기능

01Specialized monitoring for GenAI token usage and model operation duration

02Pre-configured templates for HTTP request rates and 5xx error rate ratios

03High-precision latency analysis for p95 and p99 percentiles by service

0421 GitHub stars

05Automated discovery of active metric names across varying OTel SDK versions

06Full support for local Prometheus and AWS Managed Service for Prometheus (AMP)

사용 사례

01Auditing database operation performance to identify slow queries and backend bottlenecks

02Diagnosing sudden spikes in HTTP error rates or latency across distributed microservices

03Monitoring GenAI costs and efficiency by tracking token usage per model and operation