Skip to content

Review the Threat Model

The canonical model is governance/THREAT_MODEL.md. This is a review procedure, not a duplicate inventory.

Review before adoption and whenever auth, policy, tokens, keys, revocation, audit, streams, egress, services, ports, secrets, installers, releases, or product boundaries change.

Pin source/deployed release. Gather rendered deployment, endpoints, secret flow, dependency topology, features, and targeted tests.

  1. Confirm deployment fits the documented OSS scope.
  2. Walk each boundary: untrusted input, mediation, credential, fail-closed behavior, evidence, owner.
  3. Map assets to the canonical model’s threat categories T1-T14 and run relevant negative checks.
  4. Review every canonical known limit against deployment.
  5. Record operator controls: TLS, host isolation, IdP, backup, monitoring, incident process.
  6. Reject claims relying on enterprise-only or unverified platform controls.

Retain references/results for accepted threats and owner/containment/acceptance/date for residual risks.

Disable or isolate unsupported boundaries. Report exploitable defects privately.

Apply Harden Security Posture.