sema-evalsIndependent evaluations

Hook Enforcement — 20260720T120000000Z-cursor-run1

144-trial multi-hop babel relay through Cursor CLI with a sema check wrapper gate (composer-2.5-fast): warn detects 32/32 but 25 drifted relays ship anyway; enforce halts 32/32 pre-model.

Provenance

Deviations

Condition Trials Drift detection Silent divergence Drift halted Task success False halts Malformed Hop failed
off 48 14/32 18/32 14/32 30/48 0 0 0
warn 48 32/32 0/32 7/32 23/48 0 0 0
enforce 48 32/32 0/32 32/32 48/48 0 0 0

Drift-conditioned columns use drift-injected trials as the denominator; task success uses all trials in the condition.