Skip to content
← Governance

Risk decides who has to be in the room.

Renaming a label and removing a field three other products read are not the same act, and treating them alike is why approval queues stop being read. Foundariq classifies every change by what it touches — wording and layout, behaviour, the shape of an entity, permissions, customer data. The class picks the route: automatic, human approval, or refusal, applied the same way every time rather than argued case by case. Approvals reach the person who owns that product, with the proposal, the validation result and the reviewer's objections attached.

Blanket approval trains people to click

When every change needs a signature, signatures stop meaning anything and the queue becomes a formality. Classification exists so that what reaches a human is short enough to actually read.

What it touches sets the class

Risk comes from the Blueprint itself — whether the change alters an entity's shape, a permission, a published capability or a field marked sensitive. Nothing is graded on how risky a model felt it was.

Refusal is a final answer

Some changes should never apply, whoever asks for them. Those come back as a refusal naming the rule that produced it, rather than sitting in a queue until someone senior enough waves them through.

In this capability

  • Change classification
  • Automatic low-risk apply
  • Routed human approval
  • Outright refusal path

Stop managing prompts. Start managing software strategy.

Connect your repositories. Let Foundariq understand what you have already built. Then tell it what your portfolio should build next.

Read-only to start · Your code stays in your GitHub · [email protected]