drug-interaction-checker

This environment tests an agent's ability to identify potential drug-drug interactions given a patient's current medications and a newly prescribed drug. The agent must use a drug database tool to check for interactions and provide a recommendation.

Overview

Domain: medicine Base Class: StatefulToolEnv Difficulty: medium Task: The agent must identify all significant drug-drug interactions between a patient's existing medications and a new prescription, then recommend whether the new drug should be prescribed, adjusted, or avoided.

Quickstart

Installation

uv run vf-install drug-interaction-checker

Usage

import verifiers as vf

env = vf.load_environment("drug-interaction-checker")
results = env.evaluate_sync(
    client=vf.OpenAI(),
    model="gpt-4.1-mini",
    num_examples=10,
    rollouts_per_example=1
)

Evaluation

Run an evaluation with default settings:

uv run vf-eval drug-interaction-checker

Configure model and sampling:

uv run vf-eval drug-interaction-checker \
  -m gpt-4.1-mini \
  -n 20 -r 3 -t 1024 -T 0.7

Environment Arguments

Arg	Type	Default	Description
`num_examples`	int	1000	Number of training examples
`num_eval_examples`	int	100	Number of evaluation examples
`seed`	int	42	Random seed for reproducibility

Metrics

Metric	Meaning
`reward`	Primary reward signal
`format_reward`	Format adherence reward (if applicable)

About

Generated by synthetic-rl-env-creator.

Tags: medicine