Google's open source distributed agent runtime
Google has introduced the Agent eXecutor (AX), a distributed agent runtime currently in early development. AX aims to provide a reliable and customizable environment for managing agentic loops and executions, with features like automatic recovery and event logging. The project is open for public feedback as it evolves towards a stable release.
- ▪AX is designed to coordinate agentic loops and manage executions with event logging.
- ▪It offers features such as automatic recovery from failures and support for skill, tool, and agent execution.
- ▪The project is being developed publicly to validate design decisions before its stable release.
Opening excerpt (first ~120 words) tap to expand
Agent Executor (AX) Warning🚧 This project is in early active development and WILL introduce major changes. Please note that the resumption protocols will undergo major revisions and breaking changes prior to the stable release. AX, short for Agent eXecutor, is a distributed agent runtime. It provides a runtime that coordinates agentic loops, manages executions with event logging, and communicates with both local and remote actors. AX is designed for reliability, with native support for recovery and execution resumption, even in complex distributed setups.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at GitHub.