DeepTeam is a framework to red team LLMs and AI agents.
test for local attack
sid-muralia187d27ec47e
diff --git a/b.py b/b.pynew file mode 100644index 00000000..4baf54df--- /dev/null+++ b/b.py@@ -0,0 +1,54 @@+import os+from deepeval.models import GPTModel+from deepteam.attacks.attack_simulator.attack_simulator import AttackSimulator+from deepteam.vulnerabilities import Bias, Misinformation++def test_local_baseline_attacks():+ """Test local baseline attack generation using OpenAI"""+ print("Testing local baseline attack generation with OpenAI...\n")++ # Check if OpenAI API key is set+ api_key = os.environ.get("OPENAI_API_KEY")+ if not api_key:+ api_key = input("Please enter your OpenAI API key: ")+ os.environ["OPENAI_API_KEY"] = api_key++ # Create a simulator with an OpenAI model+ openai_model = GPTModel(model_name="gpt-3.5-turbo")+ simulator = AttackSimulator(+ purpose="a financial advisor chatbot",+ max_concurrent=5,+ simulator_model=openai_model+ )++ # Define some vulnerabilities to test+ vulnerabilities = [+ Bias(),+ Misinformation()+ ]++ # Generate baseline attacks locally+ print("Generating baseline attacks locally...")+ baseline_attacks = []++ for vulnerability in vulnerabilities:+ attacks = simulator.simulate_baseline_attacks(+ attacks_per_vulnerability_type=1,+ vulnerability=vulnerability,+ ignore_errors=True+ )+ baseline_attacks.extend(attacks)++ # Print the generated attacks+ print(f"\nGenerated {len(baseline_attacks)} attacks:")+ for i, attack in enumerate(baseline_attacks):+ print(f"\n=== Attack {i+1} ===")+ print(f"Vulnerability: {attack.vulnerability}")+ print(f"Type: {attack.vulnerability_type}")+ if attack.error:+ print(f"ERROR: {attack.error}")+ else:+ print(f"Input: {attack.input}")++if __name__ == "__main__":+ test_local_baseline_attacks()\ No newline at end of file