Abstract The United States defense and critical manufacturing industrial base currently faces a systemic "Capital-Efficiency Gap," where the rigid, deterministic logic of legacy Enterprise Resource Planning (ERP) systems fails to align financial liquidity with the velocity of stochastic geopolitical signals. This operational misalignment creates profound vulnerabilities in supply chains for high-value, long-lead assets essential to national security. This Technical Report introduces the "Readiness Protocol," a decentralized multi-agent framework designed as a non-invasive intelligent overlay for legacy IT infrastructure. Synergizing Multi-Agent Reinforcement Learning (MARL) with Large Language Model (LLM) semantic processing, the architecture utilizes three autonomous agents—Inventory, Credit, and Strategic Procurement—to optimize the order-to-cash cycle under uncertainty. Monte Carlo simulations (N=1000) demonstrate that the protocol achieves a 98.9% Service Level and reduces average backorder duration by 48% compared to standard policies. Furthermore, the system demonstrates a Capital Utilization Efficiency of 5.15x, offering Small and Medium-sized Manufacturers (SMMs) a data-driven pathway to finance resilience without depleting operating cash. Source code and simulation data available at: https://github.com/trahulkumar/GovSignal-Connect