让大模型具备持续记忆与自我感知能力，可跨会话保持上下文并自检故障。
Springdrift: An Auditable Persistent Runtime for LLM Agents with Case-Based Memory, Normative Safety, and Ambient Self-Perception
- 构建带审计追踪的持久运行时，支持跨会话任务延续
- 通过案例推理与规范安全计算，实现决策可追溯与自动纠错
- 适合需要长期可靠服务的场景，如企业级智能助手
我们提出Springdrift，一个面向长期运行的大语言模型代理的持久化运行时系统。该系统集成可审计的执行底座（追加只读内存、受控进程、基于Git的恢复机制）、基于案例推理的记忆层（采用混合检索方式，对比密集余弦基线）、确定性规范安全计算（附带可审计公理链），以及每周期注入的结构化自我状态表示（感官层）来实现连续的环境感知，无需工具调用。这些特性支持传统会话制系统难以实现的行为：跨会话任务连续性、跨渠道上下文维持、端到端决策溯源重建及自我诊断。在为期23天（19个有效工作日）的单实例部署中，该代理自主诊断了自身基础设施缺陷，分类失败模式，识别出架构漏洞，并在邮件与网页渠道间保持上下文——全程未获显式指令。我们引入“人工承保人”这一概念，指代具有持续记忆、明确权限、领域自治与可追溯责任的非人类系统，区别于普通助手与自主代理，类比专业顾问关系及训练有素的工作动物的有限自主性。本研究为系统设计与部署案例报告，非基准评测。证据来自单一实例与单一操作者，旨在展示架构特性在实践中的支撑能力。系统使用Gleam语言在Erlang/OTP上实现。代码、成果与脱敏操作日志将在发表后公开于https://github.com/seamus-brady/springdrift。
原文摘要 · Abstract (English)
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.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。