← WingFoil Benchmark — campaign 33b54dd89e8b, execution 1

Tech lead: B, E, G

Harness, not model: every arm runs the same agent with the same model, claude-sonnet-5-5. Only the harness around it differs: baseline, baseline-docs, wingfoil.

Against the baseline, baseline-docs is better in 1, worse in 0 and the same in 0 of 1 comparison across category E.

Against the baseline, wingfoil is better in 1, worse in 0 and the same in 0 of 1 comparison across category E.

B and G are not covered in this campaign.

baseline baseline-docs wingfoil (wingfoil v0.3.1 · 870bd44) 100% 50% 0% baseline: 88.1%, n = 3 88.1% n = 3 baseline-docs: 78.6%, n = 3 78.6% n = 3 wingfoil (wingfoil v0.3.1 · 870bd44): 92.9%, n = 3 92.9% n = 3 E · S10@1.0 baseline: 100.0%, n = 3 100.0% n = 3 baseline-docs: 100.0%, n = 3 100.0% n = 3 wingfoil (wingfoil v0.3.1 · 870bd44): 100.0%, n = 3 100.0% n = 3 E · S8@1.0
M-Q1: the share of public hidden tests passed on the final snapshot, per arm, with the range of its runs and their n.
Categorybaselinebaseline-docswingfoil wingfoil v0.3.1 · 870bd44
B — Planning & Managementnot covered in this campaign
E — Governance & Compliance
S10@1.0
S8@1.0
S10@1.0
M-E1 not measured
M-Q1 hold-out 8/8, 8/8, 6/8 n = 3
S8@1.0
M-E1 8 n = 3 range 7–10
M-Q1 hold-out 7/7, 7/7, 7/7 n = 3
S10@1.0
M-E1 not measured
M-Q1 hold-out 6/8, 6/8, 6/8 n = 3
S8@1.0
M-E1 0 n = 3 range 0–0 −8 ▲ better beyond variance
M-Q1 hold-out 7/7, 7/7, 7/7 n = 3
S10@1.0
M-E1 not measured
M-Q1 hold-out 6/8, 8/8, 8/8 n = 3
S8@1.0
M-E1 0 n = 3 range 0–0 −8 ▲ better beyond variance
M-Q1 hold-out 6/7, 7/7, 7/7 n = 3
G — Collaborationnot covered in this campaign

Each value is the mean of its runs, with its n; a single run is preliminary, and so is a comparison unless both sides have n ≥ 3; a range is shown from n = 3. Deltas are against the baseline. How to read these results; which tools are eligible, and why.

Campaign 33b54dd89e8b, execution 1: 18 runs of claude-sonnet-5-5; 18 runs of other models, reported apart.