Scope
What’s changing
The feature, what it touches and what’s deliberately left out, taken from the spec.
Product
Every time someone on your team writes a spec, RFC or epic, Oko reads it, writes a threat model and turns the threats into requirements an engineer can pick up. Your security team reviews the result and has the final say.
How it works
Give Oko read-only access to your docs, tickets and code hosts. There’s nothing to install, and your build pipeline stays as it is.
Oko builds a picture of your services, data stores, trust boundaries and existing controls, and keeps it current as things change.
Each new or changed design gets a risk score. A ticket to tweak a settings page goes through quietly. A new way to move money gets flagged.
Oko writes up the data flows, threats and fixes. Your security team edits it, approves it or escalates it.
The requirements go into the ticket and get linked to the pull requests that implement them.
Inside a threat model
Oko sticks to the structure security people already use, so nobody has to learn a new format and every model is quick to check.
Scope
The feature, what it touches and what’s deliberately left out, taken from the spec.
Data flows
Components, data stores, external services and the trust boundaries between them.
Threats
Each piece checked against STRIDE: spoofing, tampering, repudiation, information disclosure, denial of service and elevation of privilege.
Risk
How likely each threat is and how bad it would be, given the controls you already have.
Mitigations
A testable requirement for each threat, or an accepted risk with a named owner.
History
Every approval, edit and accepted risk is versioned, so six months from now you can still see why.
Features
1.
Every PRD, RFC and epic gets a risk score and a sentence or two on why, so your team reviews the riskiest work first instead of whatever was asked about loudest.
2.
Threat models are based on your actual architecture, data flows and controls. Nobody starts from a blank whiteboard or a generic template.
3.
Short, testable requirements with acceptance criteria, written so an engineer can estimate them and get on with it.
4.
Load your security standards, compliance controls and past decisions. Oko applies them the same way in every threat model, whoever is on review that week.
5.
Link requirements to pull requests and find out when what shipped doesn’t match what was agreed.
6.
Every threat model, decision and accepted risk is versioned and searchable, which helps when an auditor asks or something goes wrong.
Sample output
Engineers don’t get a 40-page report. They get a few short, specific requirements in the ticket they’re already working on, each labeled with the threat it addresses.
Threat model by Oko · approved by AppSec
Integrations
Oko reads from these with read-only access. If you let it, it posts requirements back to your tracker, so engineers see them where they already work.
Try it free for 30 days. You don’t need a credit card, and nobody will try to book a call with you.