arXiv cs.MA· Nami Cho, Shailja Gupta·· 23 小时前AI 评分26
SCAFFOLD:面向可复用 LLM 智能体技能的证据保全式迁移
SCAFFOLD: Toward Warrant-Preserving Evidence Transport for Reusable LLM-Agent Skills
AI 导读
论文提出 SCAFFOLD 框架,主张以“有依据复用声明”(WRC)作为 LLM 智能体技能前瞻性信任的单元,将授权与版本化技能、目标上下文、证据、假设及已识别反驳条件绑定。框架进一步给出证据保全式迁移(WPET),区分证据可采性与主张真实性,并规定三值适用性及非对称上下文迁移判断。该文仅提供形式化问题陈述、可测试设计义务与报告协议,未实现验证器,也未给出实证优越性结果。
正文 · 原文
Abstract:An agent skill that succeeded yesterday may be executable today without being justified for reuse today. Tests may cover only observable contract clauses, outcome oracles may be absent, evaluations may inherit the same erroneous specification, and permissions or dependencies may change between validation and invocation. We argue that the appropriate unit of prospective trust is a Warranted Reuse Claim (WRC): a defeasible, claim-specific authorization tied to a versioned skill, target context, evidence, assumptions, and identified defeaters. The paper sharpens this position through Warrant-Preserving Evidence Transport (WPET), a proposed discipline for deciding which historical evidence remains admissible after context change. It distinguishes evidence admissibility from truth of the supported safety or effectiveness claim; states a conditional preservation result and its limits; specifies three-valued applicability and asymmetric context-transfer judgments; and illustrates how provenance dependence and policy-invalidating changes alter authorization. We explicitly relate the proposal to assurance cases, contracts, and skill-library maintenance rather than claiming novelty for their ingredients. A falsifiable longitudinal evaluation would compare WRC-guided policies with strong version-aware baselines at matched coverage, utility, and review cost. This paper contributes a formalized problem statement, testable design obligations, and reporting protocol, not an implemented verifier, empirical superiority result, or proof of universal agent safety.
| Comments: | 8 pages, 2 figures, 4 tables |
| Subjects: | Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA); Software Engineering (cs.SE) |
| Cite as: | arXiv:2611.23990 [cs.AI] |
| (or arXiv:2611.23990v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2611.23990 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Shailja Gupta [view email]
[v1]
Wed, 18 Nov 2026 07:33:15 UTC (661 KB)
来源:arXiv cs.MA · arxiv.org