Loading skill…
Skillsevaluation
Evaluation infrastructure for LLM and VLM agents.
Extensible open-source evaluation infrastructure for systematically assessing LLM/VLM agent capabilities. EMNLP 2026. Source: open-compass/AgentCompass.
Follow the README and license in the source repository. This page only restates the project’s own public description.
Clone the repository and follow its own instructions.
You will need
1. Clone
Get the project from the source URL.
git clone https://github.com/open-compass/AgentCompass.git
2. Read the source
https://github.com/open-compass/AgentCompass
Handy commands
git clone https://github.com/open-compass/AgentCompass.git
TavynSpace does not own this harness. It is open source or publicly availablecheck the source license before you use it.