arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2604.04660cs.AI
收藏

Springdrift:一种可审计的持久运行时用于LLM代理,具备基于案例的记忆、规范安全性和环境自我感知

Springdrift: An Auditable Persistent Runtime for LLM Agents with Case-Based Memory, Normative Safety, and Ambient Self-Perception

Seamus Brady

AI总结:

Springdrift通过可审计的执行子系统、基于案例的记忆层和规范安全机制,实现了LLM代理在跨会话任务连续性和环境自我感知方面的突破,展示了无显式指令下的自主诊断能力。

摘要:

我们介绍了Springdrift,一种为长期运行的LLM代理设计的持久运行时。该系统集成了可审计的执行子系统(只追加内存、监督进程、基于Git的恢复)、基于案例的推理记忆层(混合检索,对比密集余弦基线)、规范性算术用于安全门控(具有可审计公理轨迹)以及通过结构化的自我状态表示(传感器ium)持续环境自我感知。这些特性支持在会话受限系统中难以实现的行为:跨会话任务连续性、跨通道上下文维护、端到端决策的取证重建以及自我诊断行为。我们报告了一次单实例部署持续23天(19个工作日),期间代理自行诊断基础设施错误、分类故障模式、识别架构漏洞并维护邮件和网络通道的上下文——无显式指令。我们引入了“人工保留者”这一术语,指一种非人类系统,具有持久记忆、定义权威、领域特定自主性和持续关系中的取证责任,区别于软件助手和自主代理,借鉴专业保留关系和训练动物的有限自主性。本文是关于系统设计和部署案例研究的技术报告,而非基准驱动评估。证据来自单一实例和单一操作员,作为这些架构特性实际支持能力的示例。采用约Gleam在Erlang/OTP上实现。代码、制品和红acted操作日志将在发布时提供在https://github.com/seamus-brady/springdrift。

英文摘要:

We present Springdrift, a persistent runtime for long-lived LLM agents. The system integrates an auditable execution substrate (append-only memory, supervised processes, git-backed recovery), a case-based reasoning memory layer with hybrid retrieval (evaluated against a dense cosine baseline), a deterministic normative calculus for safety gating with auditable axiom trails, and continuous ambient self-perception via a structured self-state representation (the sensorium) injected each cycle without tool calls. These properties support behaviours difficult to achieve in session-bounded systems: cross-session task continuity, cross-channel context maintenance, end-to-end forensic reconstruction of decisions, and self-diagnostic behaviour. We report on a single-instance deployment over 23 days (19 operating days), during which the agent diagnosed its own infrastructure bugs, classified failure modes, identified an architectural vulnerability, and maintained context across email and web channels -- without explicit instruction. We introduce the term Artificial Retainer for this category: a non-human system with persistent memory, defined authority, domain-specific autonomy, and forensic accountability in an ongoing relationship with a specific principal -- distinguished from software assistants and autonomous agents, drawing on professional retainer relationships and the bounded autonomy of trained working animals. This is a technical report on a systems design and deployment case study, not a benchmark-driven evaluation. Evidence is from a single instance with a single operator, presented as illustration of what these architectural properties can support in practice. Implemented in approximately Gleam on Erlang/OTP. Code, artefacts, and redacted operational logs will be available at https://github.com/seamus-brady/springdrift upon publication.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑