PlanFence: Dependency-Scoped Validation to Prevent Stale-Plan Execution in Distributed LLM-Agent Memory

Forum topic · 小凯 · 2026-09-06

Summary

Distributed LLM-agent teams can read the latest shared facts and still act on an obsolete plan—a failure mode the authors call stale-plan execution. A planner may derive an action from requirement r₃, another agent commits r₄, and the executor receives r₄ without replacing the plan derived from r₃, so state freshness does not guarantee the authorizing plan remains valid. This paper introduces PlanFence, a dependency-scoped action-validation protocol: plans cite the exact public records they used, and executors validate only records that can affect the pending external action, replanning once or blocking when validation is incomplete. In 30 controlled live workflows with a post-plan revision, a freshness-only executor acts on an obsolete plan in every task, while PlanFence completes all tasks with no invalid actions. Controlled replays identify two boundary conditions: proactive synchronization yields lower coordination stalls at low change rates, while PlanFence avoids repeated update-path coordination as change rates grow and avoids validating unrelated state as the shared key space grows. The results are controlled safety and system-cost findings, not general task-accuracy improvements.

Overview

Field: AI/ML Authors: Evan Chen, Shiqiang Wang, Christopher G. Brinton Published: 2026-09-06 arXiv: 2509.00004

Key Problem: Stale-Plan Execution

Distributed LLM-agent teams can read the latest shared facts and still act on an obsolete plan. A planner may derive an action from requirement r₃, another agent may commit r₄, and an executor may receive r₄ without replacing the plan derived from r₃. The authors call this stale-plan execution: state freshness does not establish that the plan authorizing an action remains valid.

PlanFence Protocol

PlanFence is a dependency-scoped action-validation protocol:

Plans cite the exact public records they used.
An executor validates only the records that can affect the pending external action.
When validation is incomplete, the executor replans once or blocks.

Experimental Findings

In 30 controlled live workflows with a post-plan revision, a freshness-only executor acts on the obsolete plan in every task, whereas PlanFence completes all tasks with no invalid actions.
Controlled replays reveal two conditional boundaries:
Proactive synchronization produces lower coordination stalls at low change rates.
PlanFence avoids repeated update-path coordination as change rates grow, and avoids validating unrelated state as the shared key space grows.

Scope Note

These are controlled safety and system-cost results, not general task-accuracy improvements.