When risk assessment identifies an over-privileged, stale, or exposed machine identity, NHI Security can generate a Remediation Plan. To protect production infrastructure from accidental disruption, remediation operates under strict safety boundaries:
  • Plans must be validated via dry-run simulations before execution.
  • Execution against external cloud providers is gated by the deployment flag REMEDIATION_EXECUTION_ENABLED (disabled by default in self-hosted deployments).
  • Approved actions require Two-Person Integrity (dual admin approval).

1. Remediation Lifecycle


2. Generating a Plan

A remediation plan specifies the exact sequence of provider operations needed to mitigate a finding:
Example response showing planned operations and blast radius:

3. The Execution Safety Gate

Before any change is dispatched to a cloud provider, NHI Security checks the deployment configuration: Check if execution is enabled in your environment:

4. Approving and Executing Plans

Once verified, an independent Organization Admin approves execution:
Execution runs immediately upon approval. If the provider returns an error, the plan status is marked partially_failed, and an alert is recorded.

5. Audit Logging

Every plan generation, dry-run simulation, approval, and provider execution is permanently committed to the immutable audit trail:
Each record contains the operator’s identity, the target provider ARN, the parameter diff, and the cryptographic hash chaining this event to the previous record.