Why it matters
IntentGuard addresses infrastructure-as-code that is syntactically valid yet violates what a service is meant to do. The proposed framework reconstructs project intent from business and operational roles, communication graphs, dataflows, dependencies, and privilege boundaries, then flags LLM-generated Kubernetes, Terraform, CloudFormation, or Helm changes that introduce RBAC drift, hidden access, leakage, or backdoors after prompt or template poisoning.
My takeaway: Express architectural intent as testable invariants and compare every generated configuration against repository and deployment context, not only a linter rule set. Block changes that expand communication paths, privileges, or data access without explicit review, retain standard IaC scanning, and add poisoned-template and semantically plausible misconfiguration cases to regression tests.