Adnan KhanAdnanProduct & AI
San Francisco Bay Area
Get in touch

Source note3 min read

Always-OnAgents:A Survey of Persistent Memory, State, and Governance in LLMAgents

Source

arXiv

arxiv.org/abs/2606.30306v1

Read the source

Source

Plain-Language Brief

The paper 'Always-OnAgents:A Survey of Persistent Memory, State, and Governance in LLMAgents' discusses the concept of always-on agents, which are systems that have a durable state accumulated from past interactions. The authors survey various literature related to these systems through six diagnostic axes: authority, scope, mutability, provenance, recoverability, and actionability, and a lifecycle consisting of several stages like writing, validating, organizing, retrieving, acting upon, updating, forgetting, auditing, and rolling back. The survey found that the literature focuses more on accumulating and retrieving state rather than governing, recovering, or relinquishing it. To address this, they introduce the Always-On Evaluation Protocol (AOEP-v0) for evaluating these systems based on their mutation and recovery obligations.

What the Document Says

  • The paper discusses always-on agents which are systems with durable state accumulated from past interactions.
  • The authors survey literature on persistent-state systems, focusing on retrievable memories, task ledgers, permissions, credentials, commitments, provenance and audit records, shared state, trigger conditions, and externally committed effects.
  • They use six diagnostic axes: authority, scope, mutability, provenance, recoverability, and actionability to analyze the literature.
  • The lifecycle of state includes writing, validating, organizing, retrieving, acting upon, updating, forgetting, auditing, and rolling back.
  • According to their analysis, the literature focuses more on accumulating and retrieving state rather than governing, recovering, or relinquishing it.
  • To address this imbalance, they introduce the Always-On Evaluation Protocol (AOEP-v0) for evaluating these systems based on mutation and recovery obligations.
  • The agenda of the paper connects always-on agents to databases, distributed systems, formal methods, capability security, and machine unlearning.

Key Concepts Mentioned

  • always-on agents
  • persistent-state systems
  • retrievable memories
  • task ledgers
  • permissions
  • credentials
  • commitments
  • provenance and audit records
  • shared state
  • trigger conditions
  • externally committed effects
  • diagnostic axes
  • authority
  • scope
  • mutability
  • provenance
  • recoverability
  • actionability
  • lifecycle
  • writing
  • validating
  • organizing
  • retrieving
  • acting upon
  • updating
  • forgetting
  • auditing
  • rolling back
  • Always-On Evaluation Protocol (AOEP-v0)
  • databases
  • distributed systems
  • formal methods
  • capability security
  • machine unlearning

Claims Made by the Document

  • The literature focuses more on accumulating and retrieving state rather than governing, recovering, or relinquishing it.

Evidence Used

  • According to their analysis of a 435-work coded corpus, treated as a scoped map rather than an exhaustive census.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

Always-on agents are systems whose future behavior depends on durable state accumulated across earlier interactions. We treat them as persistent-state systems: the operative system includes retrievable memories, but also task ledgers, permissions, credentials, commitments, provenance and audit records, shared state, trigger conditions, and externally committed effects linked to those records. The survey reads the literature through six diagnostic axes for each state item, authority, scope, mutability, provenance, recoverability, and actionability, and through a lifecycle in which state is written, validated, organized, retrieved, acted upon, updated, forgotten, audited, and sometimes rolled back. Across a 435-work coded corpus, treated as a scoped map rather than an exhaustive census, the literature concentrates more heavily on accumulating and retrieving state than on governing, recovering, or relinquishing it. We therefore introduce the Always-On Evaluation Protocol (AOEP-v0), a pilot evaluation contract that makes these governance requirements concrete by scoring state mutation and recovery obligations rather than answer quality alone. The resulting agenda connects always-on agents to databases, distributed systems, formal methods, capability security, and machine unlearning.