Pith. sign in

← back to paper

Review history

arxiv: 2605.26563 · 3 revisions

TrajAudit: Automated Failure Diagnosis for Agentic Coding Systems

  1. 2026-08-02 CONDITIONAL MODERATE v1.3.0-alltime-deepseek novelty 6.0
    149392 ms 18123 in 14287 out 2026-08-02T13:03:48.601834+00:00
  2. 2026-07-12 CONDITIONAL LOW v1.1.0-grok45 novelty 6.0
    28909 ms 13922 in 2679 out 2026-07-12T15:55:46.109128+00:00
  3. 2026-06-29 UNVERDICTED LOW v0.9.1-grok novelty 6.0
    26157 ms 5803 in 1088 out 2026-06-29T16:10:00.088043+00:00