@premsai-pendela· Jun 23
Debugged Naga's Agent
Naga's Agent
DEEP WORK
48.0x your typical session
3.9M tokens vs typical 81.3K
5.3Mtokens
9prompts
11:21:47time
1,479lines
Claude Codeclaude-opus-4-8Markdown$150.48
mostly building
15m session running commands and testing. 3.8M tokens across 9 prompts using claude-opus-4-8.
Tokens / prompt589.2k
Cost / line$0.102
Cache hit98%
Burn rate$13.24/hr
Derived from this session's token and cost data. Not shown on the feed.
Building 100%Thinking 0%
Comments