In critical infrastructure, operational technology environments often cannot be actively scanned, yet active system feedback is needed for risk assessment and compliance management. This paper presents a non-invasive, MCP-grounded multi-agent pipeline that converts natural-language system descriptions into source-verified knowledge graphs and audit-ready artifacts in the NIST OSCAL format for continuous automated compliance management.
The architecture decouples LLM-based reasoning from deterministic knowledge retrieval against authoritative threat-intelligence sources, reducing the risk of fabricated vulnerabilities and hallucinated attack paths. In an evidence-based synthetic scenario of a water utility, the pipeline achieves 0.90 CVE recall and perfect D3FEND recall. It generates a schema-valid OSCAL System Security Plan and an OSCAL Security Assessment Report.
However, the core insight is not that grounding via MCP eliminates errors (e.g., hallucinations) entirely from the pipeline, but that it shifts errors into the first phase of asset extraction from the natural language description. Here, a single incorrectly extracted entity can lead to genuine but irrelevant CVEs in subsequent stages of the pipeline, which consumes time and resources. Nevertheless, it makes the remaining risk visible, verifiable, and suitable for a time-efficient manual review, since the infrastructure (e.g., version numbers, OS, etc.) is typically known.
Blogger's Review: This research illustrates how combining MCP with a multi-agent system effectively addresses compliance challenges in critical infrastructure. By decoupling LLM reasoning from knowledge retrieval, the potential risk of errors is significantly reduced, providing a novel approach to automated compliance management. Although there are still risks of extraction errors, the transparency it offers facilitates subsequent manual reviews.