跳到正文
arXiv cs.AI· Deepak Waghmare, Shaad Sufi·· 56 分钟前AI 评分49

AI 智能体操作可逆吗?Reversibility Classes 可逆性分级框架发布

Can You Undo It? Reversibility Classes for AI Agent Actions

AI 导读

一项 arXiv 论文提出 Reversibility Classes,为 AI 智能体工具操作定义五级可逆性词汇表,含参数敏感规则、保守默认值和先验状态捕获契约,覆盖 20 个 MCP 服务器的 536 个工具。

正文 · 原文

View PDF HTML (experimental)

Abstract:Tool-using AI agents change files, databases, access rights and public information, yet permission to call a tool does not establish whether a mistake can be undone. We present Reversibility Classes, a declared five-level vocabulary with argument-sensitive rules, conservative defaults and an explicit prior-state capture contract. The pinned implementation and evidence-linked registry cover 536 tools across 20 MCP servers. All 738 rule-case checks conform to declared expectations, while native annotation mapping differs from draft registry labels for 118 of 481 comparable tools. One AI-assisted reviewer approves 148 ratings, 75 matching the registry; this is not independent agreement. On 2,158 recorded runs, the implemented session guard gates the associated action in all 29 attributable attack-positive runs, compared with 18 for a read/write proxy, while adding nine approval points across 1,209 benign runs. Controlled file, Git, SQLite and three-browser experiments show that recovery depends on captured state, later changes, editor storage and recovery route, and expose a Git merge-state exception. The evidence supports making recovery assumptions explicit and testable. It does not establish registry-wide accuracy, reduced production harm or general recovery reliability.
Subjects: Artificial Intelligence (cs.AI)
Cite as: arXiv:2611.22075 [cs.AI]
  (or arXiv:2611.22075v1 [cs.AI] for this version)
  https://doi.org/10.48550/arXiv.2611.22075

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Shaad Sufi [view email]
[v1] Tue, 17 Nov 2026 01:55:26 UTC (135 KB)

来源:arXiv cs.AI · arxiv.org