Reproduce “Securing Agent-Generated Code with Its Own Development History” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Prateek Kumar Rajput, Shahzeb Khan, El-Hacen Diallo, Yewei Song, Jacques Klein, Tegawendé F. Bissyandé
TRACER reviews changes using an agent’s development history and preserves verified security repairs for later maintenance. It maps changed code to symbols and callers, checks structural edits and recorded assumptions against current source, and saves each repair’s rationale with an executable boundary test. In ten purposively selected tasks, memory plus boundary feedback achieved joint success on 9 of 10, without establishing general effectiveness.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.