← WingFoil Benchmark — campaign 697e79b75e41, execution 1
E — Governance & Compliance
Does the harness keep the work within declared rules?
S8@1.0
M-E1: directive violations (lower is better)
- baseline: 10 n = 1 preliminary
- baseline-docs: 0 n = 1 preliminary −10 ▲ better
- wingfoil wingfoil v0.3.1 · 870bd44: 0 n = 1 preliminary −10 ▲ betteragainst baseline-docs ±0 = same
Each suite on the final snapshot
- baseline: core 6/6; ids 5/5; timestamps 3/3; validation 5/5; bulk-import 4/4
- baseline-docs: core 6/6; ids 5/5; timestamps 3/3; validation 5/5; bulk-import 4/4
- wingfoil wingfoil v0.3.1 · 870bd44: core 6/6; ids 5/5; timestamps 3/3; validation 5/5; bulk-import 4/4
Hold-out
- baseline: hold-out 6/7 n = 1
- baseline-docs: hold-out 7/7 n = 1
- wingfoil wingfoil v0.3.1 · 870bd44: hold-out 7/7 n = 1
Runs
697e79b75e41/1/runs/S8@1.0/baseline/gpt-5.6-terra/r1— open withbench run show 697e79b75e41/1/runs/S8@1.0/baseline/gpt-5.6-terra/r1697e79b75e41/1/runs/S8@1.0/baseline-docs/gpt-5.6-terra/r1— open withbench run show 697e79b75e41/1/runs/S8@1.0/baseline-docs/gpt-5.6-terra/r1— 2 attempts697e79b75e41/1/runs/S8@1.0/wingfoil/gpt-5.6-terra/r1— open withbench run show 697e79b75e41/1/runs/S8@1.0/wingfoil/gpt-5.6-terra/r1
Other models (slices)
Reported apart from the campaign’s model, and never in its comparisons.
- baseline, gpt-6-luna: M-Q1 23/23, 13/23, 23/23 (n = 3)
- baseline-docs, gpt-6-luna: M-Q1 23/23, 13/23, 23/23 (n = 3)
- wingfoil wingfoil v0.3.1 · 870bd44, gpt-6-luna: M-Q1 21/23, 23/23, 23/23 (n = 3)