mlsanityData to 7 Oct 2026

Compile, Then Page: Executable SOP Programs and a Capability-Gated Runtime for Procedural LLM Agents

arXiv 2607.11346 ↗First posted 13 Jul 2026cs.AI cs.PLcited by 4

On the map: LLM agents and tool use → Software engineering coding agents

Abstract (arXiv)

Enterprise agents must follow long-horizon, conditional, safety-critical standard operating procedures (SOPs). We compile machine-readable SOP constraints into executable pseudo-code and run them with a program-guided (PG) stack machine that pages the active frame while an LLM performs semantic execution. A three-arm SOPBench study across six models separates representation from runtime: compiled text never significantly hurts and gains up to 16.0 points where official prose underperforms. Runtime guidance is capability-gated. Two strong models independently show positive seven-domain PG contrasts (58:19 and 75:31 discordant pairs), whereas weak models are harmed. A full-program cursor ablation (active frame first, complete program retained) recovers much of the strong-model refusal gain; selective visibility adds a smaller improvement. Paired probe and audit measurements track this divide to spontaneous state discipline rather than reconstruction ability. On Bank the three primary arms rise from 70.4 to 86.4 to 92.8, with 100% refusal correctness. Practical guidance: compile first; enable active-frame paging only after a model-level discipline check.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Placed in Software engineering coding agents, one of 16 directions in LLM agents and tool use. Builds on 13 of its 17 references that are in the map. First posted on arXiv 13 Jul 2026.

Builds on

13 of 17 references.

Closest ancestors · 1

Same direction, newest first

  1. 2025FlowAgent: Achieving Compliance and Flexibility for Workflow Agents20 Feb

Other references · 12

  1. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  2. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,383
  3. 2023MetaGPT: Meta Programming for A Multi-Agent Collaborative Frameworkcited by 1,761
  4. 2024τ-bench: A Benchmark for Tool-Agent-User Interaction in Real-World Domainscited by 1,280
  5. 2023AgentBench: Evaluating LLMs as Agentscited by 1,211
  6. 2023SGLang: Efficient Execution of Structured Language Model Programscited by 1,107
  7. 2023DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelinescited by 803
  8. 2023Efficient Guided Generation for Large Language Modelscited by 192
  9. 2024StateFlow: Enhancing LLM Task-Solving through State-Driven Workflowscited by 65
  10. 2025SOPBench: Evaluating Language Agents at Following Standard Operating Procedures and Constraintscited by 21
  11. 2025SOP-Agent: Empower General Purpose AI Agent with Domain-Specific SOPscited by 11
  12. 2026InfiAgent: An Infinite-Horizon Framework for General-Purpose Autonomous Agentscited by 8

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 4.

4 papers.

  1. 2026STAGE: Stateful Translation to Agentic Graph Execution with Policy-Scoped Context and Deterministic Controlcited by 0
  2. 2026How Strongly Should Task State Influence an LLM Agent?cited by 0
  3. 2026HEXIS: Compiling Agent Skills into Extended Finite State Machinescited by 0
  4. 2026Thinking Outside the Box: Can Language Models Rely on External Guidance Selectively?cited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Architecting Resilient LLM Agents: A Guide to Secure Plan-then-Execute Implementations
  2. 2026LLM-as-Code: Agentic Programming for Agent Harness
  3. 2026Auto: The AGI Compiler
  4. 2026Tool-Making and Self-Evolving LLM Agents in Low-Latency Systems
  5. 2026Benchmarks are Not Enough: RAMP for Runtime Assessing of Agentic Models in Production Systems
  6. 2026Skills as Verifiable Artifacts: A Trust Schema and a Biconditional Correctness Criterion for Human-in-the-Loop Agent Runtimes
  7. 2026From Agent Loops to Structured Graphs:A Scheduler-Theoretic Framework for LLM Agent Execution
  8. 2026Formal Skill: Programmable Runtime Skills for Efficient and Accurate LLM Agents