docs/testing/2026-09-30-instruction-trim/clankie-suite-claude-summary.md

clankie · claude 2.1.285 (Claude Code) · claude-sonnet-5-5

ArmPass rate95% CITokens/trial (95% CI)Wall/trial (95% CI)TrialsErrors
current10/10 (100%)72% to 100%50k (41k to 58k)7 s (5 s to 9 s)100
trimmed10/10 (100%)72% to 100%31k (23k to 40k)7 s (5 s to 10 s)100

trimmed − current, paired by case (case-clustered bootstrap):

MeasureDifference95% CIVerdict
Pass rate+0%+0% to +0%within noise (10 cases)
Pass rate: incident+0%+0% to +0%within noise (5 cases)
Pass rate: social+0%+0% to +0%within noise (5 cases)
Tokens/trial−19k−23k to −14koutside noise (10 cases)
Wall/trial+0 s−1 s to +1 swithin noise (10 cases)