Vibe Coding Discoverllm-guard
View on GitHubThe Security Toolkit for LLM Interactions
Python toolkit for scanning and sanitizing LLM inputs and outputs, with protections for prompt injection, sensitive data, secrets, and harmful content. Includes an API deployment option; the project is archived and no longer maintained.
Use Cases
Detect prompt injectionAnonymize personal informationBlock secrets and sensitive dataFilter toxic or harmful contentValidate model inputs and outputsCheck generated content for policy violations
Built With
- Language
- Python
- Frameworks
- Transformers · PyTorch · LangChain
Tags
LLM security · prompt injection · input validation · output scanning · PII detection · secret detection · content moderation · data leakage prevention · LLM firewall · archived