We all know the drill. You spend three days in a workshop, arguing about trust boundaries and data flows until someone finally draws a red line around the service account that everyone was ignoring. You export a beautiful diagram and a Confluence page gets born. Then, it fossilizes.

Six months later, during an incident, someone asks, "Is that legacy API gateway still in scope?" and the only answer is a shrug. The "living" threat model has been dead for five months and twenty-nine days.

So, the real question isn't how to *create* one, but how to keep it from becoming a historical artifact we lie to auditors about. I've seen the usual failed prescriptions:
* "Integrate it into the SDLC!" (Which means a checkbox for "threat model reviewed" that gets clicked at 4:59 PM on a Friday).
* "Assign an owner!" (Who promptly gets reassigned to put out production fires).
* "Review it quarterly!" (Be honest, when was the last quarterly task that didn't slip?).

My cynical take is that a document only stays alive if it's the source of truth for something that *actually* runs. If your automation—CI/CD, deployment pipelines, SOAR playbooks—doesn't query it or enforce its boundaries, it's just decoration.

For our internal OpenClaw agent deployment, we forced the issue by tying a snippet of the threat model (specifically, the allowed network egress paths for the collector) directly into the infrastructure-as-code. The pipeline fails if a new egress rule doesn't match an approved data flow.

```hcl
# In the terraform for the collector VPC
resource "aws_security_group_rule" "collector_egress" {
description = "Approved egress to SIEM ingest. Threat Model Ref: TM-OC-01, Data Flow D3"
cidr_blocks = [var.siem_ingest_cidr] # This var must map to a defined boundary
}
```

The document is now "alive" because it breaks the build. Painful? Yes. Effective? Also yes.

What's the least painful, most effective hook you've found to make a threat model matter *after* the pizza from the workshop is gone? Bonus points for examples that don't require a full-time priesthood to maintain.