Abstract
Existing multi-agent workspace managers emphasize task decomposition, cooperative execution, and centralized or semi-distributed orchestration. While effective in stable environments, these systems lack robustness under adversarial conditions, prolonged uncertainty, and resource contention. This paper introduces Siege-State Multi-Agent Workspace Systems (SS-MAWS), a novel framework inspired by siege warfare doctrine, in which progress is achieved not through direct execution but through gradual constraint imposition, resource isolation, and systemic pressure.
SS-MAWS reconceptualizes multi-agent coordination as a process of encirclement, containment, and controlled resolution, rather than linear task completion. The framework introduces three foundational constructs: (i) Encirclement Fields, which dynamically constrain problem spaces; (ii) Attrition-Based Execution Loops, which replace deterministic workflows with progressive convergence; and (iii) Fortification Inversion Protocols, which transform resistant problem regions into cooperative substructures. Together, these mechanisms enable resilient, long-duration multi-agent systems capable of operating under incomplete information, adversarial interference, and shifting constraints.
This approach represents a departure from execution-centric orchestration toward pressure-based computation, with significant implications for enterprise AI systems
Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 License.
Recommended Citation
Manuel-Devadoss, Johny, "Siege-State Multi-Agent Workspace Systems (SS-MAWS): Constraint-Driven, Encirclement-Based AI Coordination for Persistent Adversarial Environments", Technical Disclosure Commons, ()
https://www.tdcommons.org/dpubs_series/9898