GitHub · Rust
orchetron/secondwind
Losslessly compress the tool output your AI agent sends the model, prove every value survived, and report the exact tokens removed. Run it as a proxy, a library, or middleware.
- Stars
- 41
- Forks
- 4
- Stars gained in 24 hours
- 0
- Forks gained in 24 hours
- 0
Topics: agent, coding-agent, compression, context-engineering, context-window, multi-agent-systems, prompt-optimization, proxy, python, rust
Latest data point: Nov 21, 2026 · 41 stars · 4 forks