Claude metrics
Claude metrics for each employee, not only the API bill.
Claude Code can report sessions, token use, and cost from its own console. Employee AI usage metrics place those Claude prompts next to Cursor, Codex, and Copilot, under the person who ran them.
Fields kept for Claude
A Claude prompt keeps the employee, the agent name Claude Code, the model, and the input, output, and cached token counts. The prompt text is not required for the company report. Turns are the number of completed Claude replies stored for that person.
Anthropic’s own analytics remain the source for Claude subscription spend. This page is the cross-agent employee view: the same token columns used for every other coding agent.