AgentLab: An open-source framework for developing, testing, and benchmarking web agents on diverse tasks, designed for scalability and reproducibility.

agent agents benchmark evaluation-framework lab llm llm-agents prompting web-agents
1 Open Issue Need Help Last updated: Jul 16, 2025

Open Issues Need Help

View All on GitHub
Bring LiteLLM support about 2 months ago

AI Summary: Integrate LiteLLM support into the AgentLab framework. This involves creating a new LiteLLM model class, adding support for Anthropic parameters and other features, implementing cost tracking, writing tests, and verifying functionality by replicating existing agents and comparing outputs.

Complexity: 4/5
enhancement good first issue

AgentLab: An open-source framework for developing, testing, and benchmarking web agents on diverse tasks, designed for scalability and reproducibility.

Python
#agent#agents#benchmark#evaluation-framework#lab#llm#llm-agents#prompting#web-agents