What Is Rebuff?
Rebuff is an open-source tool for detecting and mitigating prompt injection attacks in LLM applications. It helps developers identify malicious inputs and add protective checks around model workflows.
Prompt injection is one of the most common risks in tool-using AI apps, and Rebuff gives developers a lightweight way to add defense layers.
Key Features of Rebuff
- Prompt injection detection for LLM applications.
- Mitigation patterns for unsafe or malicious inputs.
- Open-source tooling for developers building AI apps.
- Useful for chatbots, agents, RAG apps, and tool-using systems.
- Helps teams add security checks around model prompts.
Who Should Use Rebuff?
Rebuff is best for developers and security-minded AI engineers who need prompt injection defenses in LLM products.
Rebuff Pricing
Rebuff is open source. Costs depend on deployment, model usage, and any infrastructure used with it.
