cd /news/ai-agents/can-you-undo-it-reversibility-classe… · home › topics › ai-agents › article
[ARTICLE · art-171532] src=arxiv.org ↗ pub= topic=ai-agents verified=true sentiment=· neutral

Can You Undo It? Reversibility Classes for AI Agent Actions

A new arXiv paper (2611.22075v1) introduces Reversibility Classes, a declared five-level vocabulary with argument-sensitive rules and an explicit prior-state capture contract, to determine whether a tool-using AI agent's action can be undone. The pinned implementation and evidence-linked registry cover 536 tools across 20 MCP servers, with all 738 rule-case checks conforming to declared expectations, while native annotation mapping differed from draft registry labels for 118 of 481 comparable tools. On 2,158 recorded runs, the implemented session guard gated the associated action in all 29 attributable attack-positive runs versus 18 for a read/write proxy, while adding nine approval points across 1,209 benign runs; the authors state the evidence does not establish registry-wide accuracy, reduced production harm or general recovery reliability.

recommendation

Declare reversibility levels with argument-sensitive rules and capture prior state before tool calls; the session guard gated all 29 attack-positive runs but added nine approvals across 1,209 benign runs.

more in playbook: Securing agents and AI tools
by read1 min views5 publishedNov 18, 2026

arXiv:2611.22075v1 Announce Type: new 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.

── more in #ai-agents 4 stories · sorted by recency
── more on @reversibility classes 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/can-you-undo-it-reve…] indexed:0 read:1min 2026-11-18 · —