Skip to main content

​
Core Concepts

Learn the fundamental concepts of A24Z observability.

​
Sessions

A session represents a single interaction with your AI coding tool, from start to finish. Example: Opening Claude Code, writing code with AI assistance, and closing the session. What’s tracked:
  • Session duration
  • Tools used
  • Token consumption
  • Costs
  • Success rate
Learn more →

​
Tools

Tools are the individual capabilities your AI coding agent can use. Examples:
  • bash - Execute shell commands
  • view - Read files
  • str_replace - Edit files
  • search_code - Find code
Metrics:
  • Execution count
  • Success rate
  • Average duration
  • Error patterns
Learn more →

​
Metrics

A24Z tracks performance and usage metrics: Performance Metrics:
  • Tool execution time
  • Success/failure rates
  • Error rates
Usage Metrics:
  • Token consumption
  • API calls
  • Session counts
Cost Metrics:
  • Token costs
  • Cost per session
  • Cost attribution
Learn more →

​
Organizations

An organization is the top-level entity in A24Z. Key points:
  • All data belongs to an organization
  • Users can belong to multiple organizations
  • Data is strictly isolated between organizations

​
Next Steps

Sessions

Deep dive into sessions

Tools

Understanding tools

Metrics

Metrics explained

Playbooks

Role-specific guides