How does it help with debugging tool failures?

It identifies error patterns in tool results, maps tool sequences, and analyzes subagent completion rates to pinpoint exactly where complex tasks are failing.

Can I generate a summary of all activity?

Yes, the generate-report command creates a comprehensive Markdown report covering errors, performance, costs, sessions, and tool usage patterns.

Does it require a specific infrastructure?

It is designed to work with Prometheus and Loki instances where Claude Code telemetry is exported, typically accessible via a local or remote Grafana setup.

What data sources does this skill use?

It queries default OpenTelemetry (OTEL) data via Prometheus for metrics and enhanced hook-based telemetry via Loki for logs, traces, and specific events.

Can I track my Claude Code spending with this?

Yes, the skill includes specific operations to calculate token usage and estimate costs by repository, session, or specific timeframes.

Observability Analyzer

Name: Observability Analyzer
Author: adaptationio

byadaptationio

Analytics & Monitoring

Analyzes Claude Code telemetry to generate actionable insights into performance, costs, and tool usage patterns.

About

The Observability Analyzer is a specialized skill designed to provide deep visibility into Claude Code operations by querying and interpreting telemetry data from Prometheus and Loki. It enables developers to track token consumption and costs, identify performance bottlenecks in tool execution, and debug complex session lifecycles. By providing pre-built analysis commands for error grouping, subagent behavior, and context window utilization, it helps users optimize their AI workflows and maintain cost-efficient, high-performance development environments.

Key Features

In-depth session and conversation lifecycle analytics
Advanced telemetry querying using PromQL and LogQL
Comprehensive multi-dimensional reporting and error grouping
Automated performance and bottleneck identification
Detailed cost and token usage tracking per repository
0 GitHub stars

Use Cases

Analyzing context window efficiency and compaction events during long sessions
Monitoring and projecting API costs and token consumption for Claude Code
Identifying which custom tools or subagents are causing failures or latency

About

Key Features

In-depth session and conversation lifecycle analytics
Advanced telemetry querying using PromQL and LogQL
Comprehensive multi-dimensional reporting and error grouping
Automated performance and bottleneck identification
Detailed cost and token usage tracking per repository
0 GitHub stars

Use Cases

Analyzing context window efficiency and compaction events during long sessions
Monitoring and projecting API costs and token consumption for Claude Code
Identifying which custom tools or subagents are causing failures or latency