Index / Guides // 8 entries
Guides to agent session data
Reading, measuring, redacting and valuing the sessions your agents leave behind.
- 001 // Start hereHow to evaluate your AI agent sessionsA step-by-step way to measure your Claude Code and Codex history - how many sessions, how much real work, which products - and estimate what it is worth.Read +
- 002 // ValueWhat makes an AI agent session valuableBuyers of agent session data want long sessions of real work with tools, errors and recoveries. Here is what raises a session's value and what makes it worthless.Read +
- 003 // MarketWho buys AI agent session dataThe buyers of AI coding agent sessions - AI labs, data vendors and developer-tool companies - what each wants, and what they have paid in 2025 and 2026.Read +
- 004 // RightsCan you sell your AI coding sessions?Ownership, employer rights, personal data and AI providers' terms - the questions to answer before you license a Claude Code or Codex session to anyone.Read +
- 005 // SecurityHow to redact secrets and personal data from agent logsAI agent session files capture API keys, tokens, emails and customer data. How to find them in Claude Code and Codex logs, and how to mask them before sharing.Read +
- 006 // ToolsTools that read your agent session logs - ccusage and alternativesccusage, claude-code-log, claude-code-viewer, Agent Sessions, splitrail and tokscale compared - what each does with your Claude Code and Codex logs, and what it reads.Read +
- 007 // FormatWhat is inside an AI agent session fileA field-by-field look at Claude Code and Codex session files - records, messages, tool calls, results and token usage - and what each reveals about how the work went.Read +
- 008 // CompareClaude Code vs Codex CLI session logsHow Claude Code and Codex CLI differ in where they store sessions, file format, retention, compression, resume commands and training defaults - side by side.Read +