Multi-agent systems built on large language models (LLMs) are difficult to reason about. Coordination errors such as deadlocks or type-mismatched messages are often hard to detect through testing. We introduce a domain-specific language for specifying agent coordination based on message sequence charts (MSCs). This language separates message-passing structure from LLM calls, tool calls, and human control points, whose outcomes remain unpredictable.
We define the syntax and semantics of the language and present a syntax-directed projection that generates deadlock-free local agent programs from global coordination specifications. We illustrate the approach with a diagnosis consensus protocol and show how coordination properties can be established independently of LLM nondeterminism.
Additionally, we describe a runtime planning extension in which an LLM dynamically generates a coordination workflow for which the same structural guarantees apply. An open-source Python implementation of our framework is available as ZipperGen.
Blogger's Review: The proposed coordination mechanism based on message sequence charts offers a new perspective on the reliability of LLM agents. By separating uncertainty from predictable structures, it enhances the controllability of multi-agent systems, presenting significant potential for industrial applications. This approach is worth paying attention to.