Abstract
A method for coordinating task assignment across multiple AI agents in a multi-agent workflow by having each agent advertise its capabilities and constraint parameters at the start of each session, with a stateless message broker collecting those advertisements and selecting an assignment that covers the full capability set of the requested workflow at the lowest aggregate constraint cost. Each agent re-advertises per session, eliminating dependence on a persistent central skill registry. Tie-breaking uses historical completion latency self-reported by each agent. The method enables capability-based task distribution across heterogeneous AI agents without requiring any agent to hold global knowledge of its peers. Multi-agent task negotiation, stateless capability advertisement, and constraint-cost minimization are the core mechanisms of this disclosure.
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Burton, Aaron, "Multi-Agent Skill File Negotiation Protocol for Stateless Capability-Based Workflow Assignment", Technical Disclosure Commons, (April 24, 2026)
https://www.tdcommons.org/dpubs_series/9911