STAGE: Stateful Translation to Agentic Graph Execution with Policy-Scoped Context and Deterministic Control
Mengxi Luo, Changjia Chen, An Cao, Zirong Huang, Wanyi Dai
Abstract
Policy-governed agents must interpret case evidence while reliably following authorized procedures. We present STAGE, an executable-graph framework that confines model judgment to policy-scoped nodes while placing procedural control in deterministic code. We evaluate STAGE on three public policy-following benchmarks and Smart Dispute, a proprietary banking benchmark. Compared with monolithic full-policy execution, STAGE improves task success and repeated-run reliability, with its largest observed gains on the deeper workflows. On $τ^2$-bench Telecom and Smart Dispute, $\mathrm{Pass}^{3}$ improves by up to 55.0 and 65.7 percentage points, respectively. These results demonstrate the value of combining localized policy reasoning with deterministic procedural control for enterprise use.
Topics
Classified with taxonomy v2 on Sat, 5 Sept 2026.