Interface contract: Validation hypotheses

Conditionally on frozen pilot and calibration information:

1. The decoder and target are fixed before held-out validation.

2. Each setting is applied to independent repetitions of one fixed complete channel ${\Lambda }_0$.

3. Validation inputs are independent of the apparatus' fresh ancillas and hidden systems. The reference is untouched by the decoder.

4. Each measurement has a declared bounded outcome range and a calibrated relation between its mean and the ideal Choi expectation.

5. Shot counts and confidence allocations are fixed; no unaccounted optional stopping or data-dependent setting removal is used.

6. Every branch belongs to the tested channel. Abort and loss trials are not silently deleted.

7. An access audit supports the declared interface.

8. Calibration and transfer statements hold jointly except on an event of probability at most $\beta$.

A drifted, memory-dependent, or preparation-dependent apparatus is not covered unless a separate theorem supplies an appropriate sampling model.