Prompt Fuzzer is an open-source fuzzing tool designed to test and evaluate the security of system prompts in AI applications. The tool applies fuzz testing methodologies to prompt engineering, enabling developers to assess the resilience of their AI systems against potential vulnerabilities. The fuzzer operates as a fuzzing engine that generates diverse test case scenarios to evaluate system prompts. It employs dynamic LLM-based attacks and functions as a mutation-based fuzzer that creates variations of malicious prompts to test AI system defenses. The testing approach is similar to penetration testing methods used in traditional software security. The tool includes a Playground feature that allows users to iterate settings in a chat format, with each iteration representing a new fuzzing campaign. This interactive environment enables developers to test and harden their system prompts through multiple testing cycles. Prompt Fuzzer is available on GitHub and provides software testing capabilities specifically tailored for AI application security. The tool focuses on identifying weaknesses in system prompts through automated testing and quality assurance processes.