Atomix: Timely, Transactional Tool Use for Reliable Agentic Workflows

Published: 01 Mar 2026, Last Modified: 24 Apr 2026ICLR 2026 AIWILDEveryoneRevisionsCC BY 4.0
Keywords: LLM agents, transactional tool use, fault recovery, speculative execution, progress predicates, effect isolation, agentic workflows
TL;DR: Atomix wraps LLM tool calls in frontier-gated transactions so that faults trigger automatic retry and speculative branches never leak effects
Abstract: LLM agents increasingly act on external systems, yet tool effects are immediate. Under failures, speculation, or contention, losing branches can leak unintended side effects with no safe rollback. We introduce Atomix, a runtime that provides progress-aware transactional semantics for agent tool calls. Atomix tags each call with an epoch, tracks per-resource frontiers, and commits only when progress predicates indicate safety; bufferable effects can be delayed, while externalized effects are tracked and compensated on abort. Across real workloads with fault injection, transactional retry improves task success, while frontier-gated commit strengthens isolation under speculation and contention.
PDF: pdf
Email Sharing: We authorize the sharing of all author emails with Program Chairs.
Data Release: We authorize the release of our submission and author names to the public in the event of acceptance.
Submission Number: 216
Loading