TraceFix: Repairing Agent Coordination Protocols with TLA+ Counterexamples Conference: ACM Conference on AI and Agentic Systems; May 26–29, 2026; San Jose, CA, USA ACM Conference on AI and Agentic Systems (CAIS ’26), May 26–29, 2026, San Jose, CA, USA DOI: 10.1145/3786335.3813159 ISBN: 979-8-4007-2415-2/2026/05 CCS: Computing methodologies Multi-agent systems CCS: Software and its engineering Model checking CCS: Software and its engineering Software verification
Shuren Xia, Qiwei Li, Taqiya Ehsan, Jorge Ortiz · 2026
Read TraceFix: Repairing Agent Coordination Protocols with TLA+ Counterexamples Conference: ACM Conference on AI and Agentic Systems; May 26–29, 2026; San Jose, CA, USA ACM Conference on AI and Agentic Systems (CAIS ’26), May 26–29, 2026, San Jose, CA, USA DOI: 10.1145/3786335.3813159 ISBN: 979-8-4007-2415-2/2026/05 CCS: Computing methodologies Multi-agent systems CCS: Software and its engineering Model checking CCS: Software and its engineering Software verification by Shuren Xia, Qiwei Li, Taqiya Ehsan, Jorge Ortiz (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-07935.txt.
Description
Shuren Xia and three co-authors verify LLM multi-agent coordination protocols before letting them run. In TraceFix, an agent drafts a coordination protocol as a structured representation from the task description, and generates formal coordination logic. It then repairs the logic against counterexamples from the TLA+ model checker until verification passes. Verified parts become per-agent system prompts, guarded at runtime by a monitor that rejects off-protocol moves.
Across forty-eight tasks in sixteen scenario families, every task reached full verification, most on the first attempt and none needing more than four repairs. In thousands of runtime runs, monitored execution with verified protocols achieved the highest task completion, and an ablation showed verification cut deadlock and livelock rates by more than half.
Catalog record
| CREATOR | Shuren Xia, Qiwei Li, Taqiya Ehsan, Jorge Ortiz |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 72 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.07935