Search papers, labs, and topics across Lattice.
The paper introduces CBCL, a novel agent communication language (ACL) designed for safe extensibility by restricting all messages and runtime language extensions to the deterministic context-free language (DCFL) class. CBCL allows agents to define and share domain-specific language extensions as first-class messages while enforcing safety invariants to prevent unbounded expansion and resource exhaustion. The authors formalize CBCL and its safety properties in Lean 4, implement a reference parser and dialect engine in Rust, and extract a verified parser binary, demonstrating the feasibility of provably safe homoiconic protocol design.
Autonomous agents can safely extend their communication capabilities with provable guarantees, thanks to a novel language design that bounds expressiveness.
Agent communication languages (ACLs) enable heterogeneous agents to share knowledge and coordinate across diverse domains. This diversity demands extensibility, but expressive extension mechanisms can push the input language beyond the complexity classes where full validation is tractable. We present CBCL (Common Business Communication Language), an agent communication language that constrains all messages, including runtime language extensions, to the deterministic context-free language (DCFL) class. CBCL allows agents to define, transmit, and adopt domain-specific"dialect"extensions as first-class messages; three safety invariants (R1--R3), machine-checked in Lean 4 and enforced in a Rust reference implementation, prevent unbounded expansion, applying declared resource limits, and preserving core vocabulary. We formalize the language and its safety properties in Lean 4, implement a reference parser and dialect engine in Rust with property-based and differential tests, and extract a verified parser binary. Our results demonstrate that homoiconic protocol design, where extension definitions share the same representation as ordinary messages, can be made provably safe. As autonomous agents increasingly extend their own communication capabilities, formally bounding what they can express to each other is a precondition for oversight.
