Reproduce “Agent-First Tool API: A Semantic Interface Paradigm for Enterprise AI Agent Systems” 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
Kai Pan, Rong Hou
Agent-First redesigns enterprise tool interfaces to help language-model agents act on business systems. Its six-verb protocol supports descriptive search, disambiguation, previews, execution, verification, and recovery, while a normalized response contract and governance checks structure results and mutations. In a 50-task test, it achieved 88% success versus 64% for CRUD+ReAct, but this bundled comparison covers one production system.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.