AI coding agents can already autonomously execute most of an experimental high energy physics analysis pipeline. We demonstrate this with Just Furnish Context (JFC), a framework combining coding agents, literature-based knowledge retrieval, specification documents encoding HEP conventions, and multi-agent review. From a short physics prompt, Claude Code carries out event selection, background estimation, systematic uncertainty evaluation, statistical inference, and paper drafting, with no human intervention beyond an unblinding gate. We validate the framework on ALEPH, DELPHI, and CMS open data, performing electroweak, QCD, and Higgs measurements; the resulting reports are, at cursory expert review, indistinguishable from those of a junior graduate student. Rather than replacing physicists, such tools offload the repetitive technical burden of analysis development, freeing researchers for physics insight, method development, and validation.