Show HN: Teleport-env – <500ms stateful rollbacks for AI agents via CRIU
Teleport-Env is a new sandbox designed for coding agents that allows for ultra-fast stateful rollbacks. It utilizes a unique architecture to achieve recovery times of less than 500 milliseconds, significantly faster than traditional Docker containers. This innovation is particularly beneficial for high-throughput Monte Carlo Tree Search and reinforcement learning applications.
- ▪Teleport-Env provides a rollback mechanism inspired by DeltaBox, enabling sub-millisecond recovery times.
- ▪The system uses OverlayFS and CRIU to manage the agent's workspace and memory state efficiently.
- ▪It is designed to run natively on Linux environments, bypassing limitations found in Windows and macOS.
Opening excerpt (first ~120 words) tap to expand
Teleport-Env An ultra-fast, OS-level snapshot and rollback sandbox designed for autonomous coding agents, Monte Carlo Tree Search (MCTS), and reinforcement learning. 1. The Problem Coding agents need environments to test generated bash commands and scripts. However, standard Docker containers take 3 to 5 seconds to restart when an agent inevitably destroys the filesystem or corrupts a background process. This high latency makes it impossible to run high-throughput MCTS search loops or deep reinforcement learning across thousands of branches. 2. Inspiration Teleport-Env is directly inspired by DeltaBox ("DeltaBox: A Fast File-System Rollback Facility for Destructive Agentic Testing"). DeltaBox proposed a sub-millisecond rollback mechanism using a customized kernel module.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at GitHub.