How does session management work?

The skill automatically groups related operations into sessions. A session is created on the first operation and remains active until there is one hour of inactivity.

Can I track API costs with this skill?

Yes, the logger tracks token counts for each operation and provides summary methods to calculate total usage and estimated costs over specific timeframes.

Is this skill suitable for debugging?

Absolutely. It records operation success/failure status and duration, and includes specialized methods to retrieve recent errors for rapid troubleshooting.

What format are the logs stored in?

Logs are stored in structured JSONL (JSON Lines) format, making them easy to query with command-line tools like jq or import into external analytics platforms.

Where are the log files saved?

By default, log files are stored locally on your machine at ~/.claude/leyline/usage/{service}.jsonl, allowing for private and persistent data storage.

Usage Logging

Name: Usage Logging
Author: athola

byathola

•

Analytics & Monitoring

Tracks AI operations and costs with session-aware JSONL logging for audit trails and performance analytics.

About

The Usage Logging skill provides a robust infrastructure for monitoring Claude Code interactions by implementing structured JSONL logging with automatic session management. It enables developers to maintain detailed audit trails, track token usage costs across multiple operations, and build comprehensive usage analytics. By grouping related operations into sessions and providing built-in query methods, it simplifies the process of debugging complex AI workflows and managing resource consumption in production-grade environments.

Key Features

Automatic session management with one-hour inactivity timeouts
10 GitHub stars
Comprehensive metrics including token counts, success rates, and duration
Customizable metadata support for domain-specific operation tracking
Structured JSONL output for easy parsing and CLI tool integration
Built-in query methods for recent errors and usage summaries

Use Cases

Monitoring and forecasting API costs and token consumption
Auditing AI operations for compliance and security oversight
Debugging complex workflows using historical operation data and session IDs

About

Key Features

Automatic session management with one-hour inactivity timeouts
10 GitHub stars
Comprehensive metrics including token counts, success rates, and duration
Customizable metadata support for domain-specific operation tracking
Structured JSONL output for easy parsing and CLI tool integration
Built-in query methods for recent errors and usage summaries

Use Cases

Monitoring and forecasting API costs and token consumption
Auditing AI operations for compliance and security oversight
Debugging complex workflows using historical operation data and session IDs