Trace Integrity for LLM Data Agents: Auditing Structured-Data Computation Beyond Answer Accuracy in Real-World Systems.

arXiv:2608.26036v2 Announce Type: replace Abstract: Large language model (LLM) data agents answer structured-data questions by generating SQL, invoking tools, executing queries, and summarizing results. In operational settings, reliability depends on both repo