ATC-BENCH · A LIVING BENCHMARK REPORT ReportReplaysIncidentsParticipateGitHub

Appendix A · the docket

Incident reports

Every cardinal violation auto-generates an NTSB-style incident report: a synopsis, a sequence of events, and a probable-cause chain traced through event provenance. Nothing is written by hand. The first fully-published report is below; reports for the remaining pilot-campaign violations publish with the v0.2 data release.

Docket. Cardinal violations, pilot campaign, data version runs/pilot.

ReportModelPositionViolationS_rawStatus
AB-3605-01sonnet-5TWRNEGLECT0.798published →
Reports for the remaining pilot-campaign cardinal violations publish with v0.2.
ATC-Bench · 2026 · Apache-2.0 ← Back to the report