Is real-time monitoring supported?

Yes, all imported dashboards include auto-refresh capabilities (ranging from 30 seconds to 5 minutes) and time range selectors for real-time observability.

What dashboards are included by default?

The skill includes five pre-built dashboards: Claude Code Overview, Tool Performance Matrix, Cost Analysis, Error Tracking, and Session Analysis.

Can I create my own custom visualizations?

Yes, you can use the 'create-custom' command to build specific dashboards based on templates for tool-specific metrics, costs, errors, or sessions.

Do I need a running Grafana instance?

Yes, this skill requires access to a Grafana API (typically running on localhost:3000) to import, export, and manage the dashboard configurations.

How do I backup my existing Claude Code dashboards?

Use the 'export-dashboards' command to save all your current dashboard configurations as JSON files in the .observability/backups directory.

Claude Code Observability Dashboards

Name: Claude Code Observability Dashboards
Author: adaptationio

byadaptationio

0•

分析と監視

Deploys and manages comprehensive Grafana dashboards for monitoring Claude Code performance, costs, and session health.

The Observability Dashboard Creator streamlines the monitoring of Claude Code by automating the setup and management of specialized Grafana visualizations. It provides developers with immediate insights into token usage, API costs, tool execution performance, and error patterns through pre-configured dashboards. Whether you are troubleshooting failed tool calls, tracking project budgets, or analyzing conversation depth, this skill simplifies the complex process of infrastructure monitoring into a series of automated commands, ensuring your AI development environment remains transparent and optimized.

主な機能

01Performance matrix for monitoring tool latency and success/failure rates

02Automated import of 5 pre-built Grafana dashboards for immediate visibility

03Template-based custom dashboard creation for specific monitoring needs

04Real-time tracking of token consumption and granular API cost analysis

05Detailed error pattern detection and session history logging

060 GitHub stars

ユースケース

01Setting up a centralized monitoring hub after deploying a Claude Code observability stack

02Analyzing the cost-efficiency and token usage of long-running AI development sessions

03Identifying performance bottlenecks and high-failure rates in custom tool implementations

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add adaptationio/skrillz observability-dashboard-creator

For use in Claude.ai and ChatGPT

Download Skill