@premsai-pendela· Jun 21
Debugged Naga's Agent
Naga's Agent
DEEP WORK
First session with opus-4-8
First build using this model
4.5Mtokens
6prompts
42:25:00time
2,810lines
Claude Codeclaude-opus-4-8Python$167.33
mostly building
4h 35m session running commands and testing. 2.3M tokens across 4 prompts using claude-opus-4-8.
Tokens / prompt755.6k
Cost / line$0.060
Cache hit99%
Burn rate$3.95/hr
Derived from this session's token and cost data. Not shown on the feed.
Building 96%Verifying 2%Thinking 3%
Comments