evaluate-agent-runs

Agent Runs - Evaluation

In this guide, we build a simple agent which can use two tools.
We run the agent against a couple of user queries and monitor the agent runs via Literal AI.
Finally, we retrieve the agent runs and check the number of tools they called.

Setup

To install dependencies, run:

pip install -r requirements.txt

Create and set your Literal AI and OpenAI API keys in .env:

cp .env.example .env