Human–Machine Relations Lab
DEContact

Evidence for assurance

Connect tested behaviour and action records to an assessment someone else can inspect.

Read more

Platform

Identities, mandates and gateway checks are the platform layer. The supplied deck labels this work as built with receipts in progress; individual integrations require their own coverage evidence.

Attestation

The next offering combines sandbox tests and receipts for a workflow review. This is an intended evidence product, not a certification or a statement that an insurer has accepted the evidence.

Risk calculator

The calculator is in design. The intended inputs are a questionnaire, test results and mandate limits. Intended outputs are a risk tier and an inspectable data sheet. The visualisation below explains the process; it does not calculate a price.

Risk Level Calculator

How evidence becomes
a risk profile.

The calculator is intended to bring together a questionnaire, test results and mandate limits. This interactive view explains the path from inputs to assessment.

Explore the process
01

Start with the workflow.

↳

Questionnaire

Task, access and possible consequences.

↳

Sandbox & receipts

Observed behaviour and open findings.

↳

Mandate

Allowed actions, counterparties and limits.

02

Assess evidence in context.

What can the agent affect?

Which limits apply?

Were those limits tested?

Which evidence is missing?

03

An inspectable risk profile.

Risk tier

Data sheet

Findings & limitations

An assessment remains bound to the tested workflow and its configuration.

Process visualisation · Calculator in design · No calculated risk tier or price.