agentdojo 0.1.17

A Dynamic Environment to Evaluate Attacks and Defenses for LLM Agents

AgentDojo: Benchmarking the Capabilities and Adversarial Robustness of LLM Agents

Edoardo Debenedetti¹, Jie Zhang¹, Mislav Balunović^1,2, Luca Beurer-Kellner^1,2, Marc Fischer^1,2, Florian Tramèr¹

¹ETH Zurich and ²Invariant Labs

Read our paper here.

Quickstart

pip install agentdojo

[!IMPORTANT] Note that the API of the package is still under development and might change in the future.

If you want to use the prompt injection detector, you need to install the transformers extra:

pip install "agentdojo[transformers]"

Running the benchmark

The benchmark can be run with the benchmark script. Documentation on how to use the script can be obtained with the --help flag.

For example, to run the workspace suite on the tasks 0 and 1, with gpt-4o-2024-05-13 as the LLM, the tool filter as a defense, and the attack with tool knowlege, run the following command:

python -m agentdojo.scripts.benchmark -s workspace -ut user_task_0 \
    -ut user_task_1 --model gpt-4o-2024-05-13 \
    --defense tool_filter --attack tool_knowledge

To run the above, but on all suites and tasks, run the following:

python -m agentdojo.scripts.benchmark --model gpt-4o-2024-05-13 \
    --defense tool_filter --attack tool_knowledge

Documentation of the Dojo

Take a look at our documentation.

Development set-up

Take a look at the development set-up docs.

Citing

If you use AgentDojo in your research, please consider citing our paper:

@misc{debenedetti2024agentdojo,
      title={AgentDojo: A Dynamic Environment to Evaluate Attacks and Defenses for LLM Agents},
      author={Edoardo Debenedetti and Jie Zhang and Mislav Balunović and Luca Beurer-Kellner and Marc Fischer and Florian Tramèr},
      year={2024},
      eprint={2406.13352},
      archivePrefix={arXiv},
      primaryClass={cs.CR}
}