mlsanityData to 7 Oct 2026

STAGE: Stateful Translation to Agentic Graph Execution with Policy-Scoped Context and Deterministic Control

arXiv 2608.22538 ↗First posted 23 Aug 2026cs.AIcited by 0

On the map: LLM agents and tool use → Agent threats, guardrails and governance

Abstract (arXiv)

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 τ2-bench Telecom and Smart Dispute, Pass3 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.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Agent threats, guardrails and governance, one of 16 directions in LLM agents and tool use. Builds on 19 of its 22 references that are in the map. First posted on arXiv 23 Aug 2026.

Builds on

19 of 22 references.

Closest ancestors · 4

Same direction, newest first

  1. 2026PolicyGuard: A Dialogue-Grounded Sub-Agent Verifier for Policy Adherence in LLM Agents28 Jun
  2. 2026Solver-Aided Verification of Policy Compliance in Tool-Augmented LLM Agents20 Mar
  3. 2025Towards Enforcing Company Policy Adherence in Agentic Workflows22 Jul
  4. 2025AgentSpec: Customizable Runtime Enforcement for Safe and Reliable LLM Agents24 Mar

Other references · 15

  1. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,383
  2. 2025τ2-Bench: Evaluating Conversational Agents in a Dual-Control Environmentcited by 558
  3. 2025Plan-and-Act: Improving Planning of Agents for Long-Horizon Taskscited by 190
  4. 2026SkillX: Automatically Constructing Skill Knowledge Bases for Agentscited by 69
  5. 2026Adapting the Interface, Not the Model: Runtime Harness Adaptation for Deterministic LLM Agentscited by 22
  6. 2025SOP-Bench: Complex Industrial SOPs for Evaluating LLM Agentscited by 19
  7. 2026Beyond IVR: Benchmarking Customer Support LLM Agents for Business-Adherencecited by 11
  8. 2026On Information Self-Locking in Reinforcement Learning for Active Reasoning of LLM agentscited by 6
  9. 2026AgentLTL: A Trace-Verification Framework for Measuring, Enforcing, and Training Procedural Compliance in Tool-Using LLM Agentscited by 4
  10. 2026Compile, Then Page: Executable SOP Programs and a Capability-Gated Runtime for Procedural LLM Agentscited by 4
  11. 2025Matrix: Peer-to-Peer Multi-Agent Synthetic Data Generation Frameworkcited by 3
  12. 2026From Atomic Actions to Standard Operating Procedures: Iterative Tool Optimization for Self-Evolving LLM Agentscited by 3
3 more
  1. 2026Declarative Skills for AI Agents in Knowledge-Grounded Tool-Use Workflowscited by 2
  2. 2026PolicyGuide: From Guarding One Action to Guiding the Whole Workflow for Policy-Compliant LLM Agentscited by 2
  3. 2026COVENANT: Natural-Language Workflow Compilation for Aligned Agent Executioncited by 1

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026GraphBit: A Graph-based Agentic Framework for Non-Linear Agent Orchestration
  2. 2026Runtime Governance for AI Agents: Policies on Paths
  3. 2026DNative-Twin: Decision Graphs and Digital Twins for Reconstructable Agentic Decisions
  4. 2026Formal Policy Enforcement for Real-World Agentic Systems
  5. 2026AgentFlow: A Flow-Centric Policy Language and Framework for Securing LLM Agent Systems
  6. 2026Policy-Invisible Violations in LLM-Based Agents
  7. 2026PolicyBank: Evolving Policy Understanding for LLM Agents
  8. 2026Graph Is the Verifier: Agentic Reinforcement Learning for Interprocedural Vulnerability Detection