Debugged fielddaycommander
2.7M tokens vs typical 207.2K
9m session running commands and testing. 206K tokens across 5 prompts using claude-opus-5-5.
Derived from this session's token and cost data. Not shown on the feed.
Comments