Metadata-Version: 2.5
Name: roboharm
Version: 0.0.1
Summary: RoboHarm: BenchFlow robotic agent safety benchmark, an AgentHarm-style paired harmful/benign evaluation (name reserved; code lives in the private robo-use-team repository until release)
Project-URL: Homepage, https://github.com/benchflow-ai
Author-email: Xiangyi Li <xiangyi@benchflow.ai>
License-Expression: Apache-2.0
Keywords: benchflow,benchmark,manipulation,robotics,safety
Classifier: Development Status :: 1 - Planning
Classifier: Programming Language :: Python :: 3
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Requires-Python: >=3.10
Description-Content-Type: text/markdown

# roboharm

RoboHarm: BenchFlow robotic agent safety benchmark, an AgentHarm-style paired harmful/benign evaluation (name reserved; code lives in the private robo-use-team repository until release)

This is a placeholder release that reserves the package name for the BenchFlow team. It contains no functionality yet. The initial public release will replace it; until then, the source of record is the private `benchflow-ai/robo-use-team` repository.
