Saved in:
Bibliographic Details
Main Authors: Zhang, Shuyu, Shi, Yaqi, Wang, Lu
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.29313
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911726433206272
author Zhang, Shuyu
Shi, Yaqi
Wang, Lu
author_facet Zhang, Shuyu
Shi, Yaqi
Wang, Lu
contents LLM multi-agent systems often coordinate through natural-language dialogue or loosely structured shared memory, making intermediate state difficult to validate, attribute, and audit. We introduce PatchBoard, a schema-grounded collaboration architecture that replaces inter-agent dialogue with validated JSON Patch mutations over a shared structured state. An Architect agent constructs a task-specific schema and workflow rules, while a deterministic kernel validates each proposed state mutation against schema constraints, role-specific write contracts, and runtime invariants before committing it transactionally. On 630 matched ALFWorld episodes, PatchBoard achieves an 84.6% success rate, compared with 30.8% for LangGraph and 61.6% for Flock, while reducing tokens per successful task to 45.5k, compared with 368.3k and 64.2k, respectively.
format Preprint
id arxiv_https___arxiv_org_abs_2605_29313
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle PatchBoard: Schema-Grounded State Mutation for Reliable and Auditable LLM Multi-Agent Collaboration
Zhang, Shuyu
Shi, Yaqi
Wang, Lu
Computation and Language
LLM multi-agent systems often coordinate through natural-language dialogue or loosely structured shared memory, making intermediate state difficult to validate, attribute, and audit. We introduce PatchBoard, a schema-grounded collaboration architecture that replaces inter-agent dialogue with validated JSON Patch mutations over a shared structured state. An Architect agent constructs a task-specific schema and workflow rules, while a deterministic kernel validates each proposed state mutation against schema constraints, role-specific write contracts, and runtime invariants before committing it transactionally. On 630 matched ALFWorld episodes, PatchBoard achieves an 84.6% success rate, compared with 30.8% for LangGraph and 61.6% for Flock, while reducing tokens per successful task to 45.5k, compared with 368.3k and 64.2k, respectively.
title PatchBoard: Schema-Grounded State Mutation for Reliable and Auditable LLM Multi-Agent Collaboration
topic Computation and Language
url https://arxiv.org/abs/2605.29313