cff-version: 1.2.0
message: "If you use OrchestrateBench, please cite this repository using the metadata below."
title: "OrchestrateBench"
type: software
version: 0.1.0
date-released: 2026-06-20
abstract: >-
  OrchestrateBench is a research benchmark for evaluating LLM orchestration
  routing, failure recovery, and cascade propagation in multi-agent systems.
authors:
  - family-names: Chen
    given-names: Yidian
  - family-names: Gu
    given-names: Yingzi
  - family-names: Vidra
    given-names: Natan
  - family-names: Setty
    given-names: Spurthi
  - family-names: Zheng
    given-names: Sharon
repository-code: "https://github.com/anote-ai/Research-OrchestrateBench"
url: "https://github.com/anote-ai/Research-OrchestrateBench"
keywords:
  - llm orchestration
  - multi-agent systems
  - benchmarking
  - failure recovery
  - cascade propagation