Fresh Memory, Stale Plans: Dependency-Scoped Validation for Distributed LLM-Agent Memory
On the map: LLM agents and tool use → Agent memory systems
Abstract (arXiv)
A large language model (LLM) agent that inherits a plan through shared memory can hold the latest requirement yet act on a plan derived from an older one: fresh memory, stale plan. Freshness checks miss this failure because they compare local copies with current state (observation currency) rather than the inputs the plan was derived from (derivation currency). Planfence makes derivation currency checkable after a handoff. Stored plans carry exact links to their recorded inputs; before a protected action, Planfence follows those links to an action-specific dependency frontier, asks each input's owner for its current head, refreshes what changed, and allows one replan before blocking. Application code supplies the links and declares the scope; no shared memory service is required. Holding the native S-Bus validator fixed, supplying inherited input versions raises detected handoff conflicts from 0/30 to 30/30: retained evidence is the missing ingredient. In 30 live five-agent workflows with a revision inserted after planning, a freshness-only executor acts on the stale plan every time, whereas Planfence, like a centralized-lineage baseline that requires a shared store, completes all 30 correctly. In matched replay under emulated LTE traces, Planfence's stall stays within 143-237ms per action across a 64× range of update rates while per-update synchronization grows from 52 to 1196ms; synchronization is cheaper only at the lowest tested rates. Scoping queries to the declared dependencies holds traffic at 8.1KiB per action, a tenth of all-key validation at 128 keys; at full scope the two tie.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in Agent memory systems, one of 16 directions in LLM agents and tool use. Builds on 16 of its 32 references that are in the map. First posted on arXiv 3 Sep 2026.
Builds on
16 of 32 references.
Closest ancestors · 8
Same direction, newest first
- 2026When Memory Updates but Behavior Does Not: Repairing Implicit Stale Dependencies in Personalized Agent Responses3 Aug
- 2026MemTX: Transactional Belief Commit for Stateful Agent Memory27 Jul
- 2026S-Bus: Automatic Read-Set Reconstruction for Multi-Agent LLM State Coordination16 May
- 2026Multi-Agent Memory from a Computer Architecture Perspective: Visions and Challenges Ahead9 Mar
- 2025MIRIX: Multi-Agent Memory System for LLM-Based Agents10 Jul
- 2025Collaborative Memory: Multi-User Memory Sharing in LLM Agents with Dynamic Access Control23 May
- 2025Mem0: Building Production-Ready AI Agents with Scalable Long-Term Memory28 Apr
- 2025A-MEM: Agentic Memory for LLM Agents17 Feb
Other references · 8
- 2023MetaGPT: Meta Programming for A Multi-Agent Collaborative Frameworkcited by 1,761
- 2023CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Societycited by 1,348
- 2023MemGPT: Towards LLMs as Operating Systemscited by 1,295
- 2025Multi-agent Architecture Search via Agentic Supernetcited by 164
- 2026Temporary Authority, Permanent Effects: Commit-Time Authorization for LLM Agentscited by 17
- 2025Beyond Connectivity: An Open Architecture for AI-RAN Convergence in 6Gcited by 10
- 2026SwarmWorld: Stigmergic technological evolution in societies of language-model agentscited by 3
- 2026SyncPlan: Long-Horizon LLM Coordination with Explicit Synchronization and Adaptive Correctioncited by 1
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026Plans Don't Persist: Why Context Management Is Load Bearing for LLM Agents
- 2025ALAS: Transactional and Dynamic Multi-Agent LLM Planning
- 2026Forgetting Without Restarting: Execution-State Unlearning for Stateful LLM Agents
- 2026TomasuLLM: Out-of-Order Speculative Execution for LLM Agents
- 2025SagaLLM: Context Management, Validation, and Transaction Guarantees for Multi-Agent LLM Planning
- 2026Coupling Planning with Episodic Memory in LLM Agents for Software Issue Resolution
- 2026Auditing Action Settlement in LLM Agent Environments: Order, Progress, and Replay
- 2026MemGuard: Persisting Verifier Signals for LLM-Agent Memory Governance