Validation instrument
Applicant study recorder
A local, anonymous recorder for the prewritten three-task protocol. It calculates the release gate exactly as specified; it cannot turn simulated runs into human evidence.
Demonstration data
Before each session
Use fictional records, do not coach, counterbalance interface order, and record failures as failures. Open the neutral baseline or open the Career ID task. See the published evidence boundary.
Loading local study records…