What file format does this skill require?

It is specifically designed to parse the .jsonl files automatically generated by Claude in the .claude-trace directory.

Does this skill help reduce AI costs?

Yes, by identifying high token consumption patterns and inefficient tool calls, you can optimize your workflows to be significantly more cost-effective.

Can I compare multiple sessions at once?

Yes, the 'compare' action allows you to aggregate metrics across several trace files to identify performance trends and usage anomalies over time.

Is this useful for debugging tool failures?

Absolutely. It categorizes errors like rate limits, timeouts, and execution failures, providing recommendations on how to resolve recurring technical bottlenecks.

How does this differ from the /transcripts command?

While /transcripts shows the conversation content, this skill analyzes the API-level JSONL data for technical metrics like latency, token usage, and specific tool execution errors.

Claude Trace Analytics

Name: Claude Trace Analytics
Author: rysweet

byrysweet

•

Analytics & Monitoring

Analyzes Claude Code trace files to provide actionable insights into session health, token consumption, and tool performance.

About

Claude Trace Analytics is a specialized diagnostic skill designed to decode and interpret the underlying JSONL trace data generated during Claude Code sessions. Unlike standard transcripts that focus on conversation history, this skill drills into API-level metrics, allowing developers to diagnose latency issues, identify recurring error patterns, and optimize token efficiency. By providing deep visibility into tool execution frequencies and model interactions, it helps users refine their agentic workflows, troubleshoot failures, and manage operational costs effectively.

Key Features

Automated error pattern detection and categorization
17 GitHub stars
Comprehensive token usage and efficiency reporting
Tool usage optimization and bottleneck identification
JSONL trace file parsing and automated health scoring
Comparative performance analysis across multiple sessions

Use Cases

Debugging slow or failing Claude Code sessions to find technical root causes
Auditing token expenditure and optimizing prompt efficiency to reduce costs
Analyzing tool execution patterns to identify parallelization opportunities

About

Key Features

Automated error pattern detection and categorization
17 GitHub stars
Comprehensive token usage and efficiency reporting
Tool usage optimization and bottleneck identification
JSONL trace file parsing and automated health scoring
Comparative performance analysis across multiple sessions

Use Cases

Debugging slow or failing Claude Code sessions to find technical root causes
Auditing token expenditure and optimizing prompt efficiency to reduce costs
Analyzing tool execution patterns to identify parallelization opportunities