Checking who's building…
Debugged Naga's Agent
5.5M tokens vs typical 74.0K
22m session running commands and testing. 5.5M tokens across 16 prompts using claude-haiku-4-5-20251001.
Derived from this session's token and cost data. Not shown on the feed.
Comments