LLM Guard Rails - Comprehensive Security Toolkit for LLM Interactions

About: This is an actively maintained fork of the original LLM Guard project by Protect AI. We maintain enhancements and improvements beyond the original project, ensuring continuous updates and feature development.

LLM Guard Rails is a comprehensive security toolkit designed to fortify the safety and security of Large Language Model (LLM) interactions. It provides robust protection against various attack vectors and malicious inputs with actively maintained enhancements.

Documentation | PyPI Package | Original Project | Getting Started

What is LLM Guard?

By offering sanitization, detection of harmful language, prevention of data leakage, and resistance against prompt injection attacks, LLM-Guard ensures that your interactions with LLMs remain safe and secure.

Installation

Begin your journey with LLM Guard Rails by downloading the package:

pip install llm-guardrails-maintained

Or install from source:

git clone https://github.com/Nikhilsinghbora/llm-guardrails.git
cd llm-guardrails
pip install -e .

Getting Started

Requirements & Notes:

• Python: 3.10, 3.11, 3.12, 3.13, or 3.14 (python --version to verify)
• Production-Ready: Designed for easy integration in production environments with continuous improvements
• Minimal Dependencies: Base functionality requires minimal libraries; advanced features auto-install as needed
• Troubleshooting: If you encounter installation issues, upgrade pip: python -m pip install --upgrade pip

Quick Examples:

• OpenAI Integration - Secure ChatGPT interactions
• LangChain Integration - LLM protection with LangChain
• API Deployment - Deploy as a service
• More Examples - Amazon Bedrock, Google Gemini, and streaming examples

Supported scanners

Prompt scanners

• Anonymize
• ATRScanner ⭐ NEW
• BanCode
• BanCompetitors
• BanSubstrings
• BanTopics
• Code
• Gibberish
• InvisibleText
• Language
• PromptInjection
• Regex
• Secrets
• Sentiment
• TokenLimit
• Toxicity

Output scanners

• ATRScanner ⭐ NEW
• BanCode
• BanCompetitors
• BanSubstrings
• BanTopics
• Bias
• Code
• Deanonymize
• JSON
• Language
• LanguageSame
• MaliciousURLs
• NoRefusal
• ReadingTime
• FactualConsistency
• Gibberish
• Regex
• Relevance
• Sensitive
• Sentiment
• Toxicity
• URLReachability

Community, Contributing, Docs & Support

LLM Guard Rails is an open source solution maintained as an active fork of the original LLM Guard project. We are committed to a transparent development process and highly appreciate any contributions. Whether you are helping us fix bugs, propose new features, improve our documentation or spread the word, we would love to have you as part of our community.

How We Differ

• Active Development: We maintain and release updates more frequently than the original project
• Feature Requests Welcome: We actively review and implement community feature requests
• Responsive Maintenance: Issues and PRs receive timely reviews and updates

Recent Improvements & Updates

Security & Bug Fixes:

• ✅ Fixed CVE-2026-26007: Updated presidio-anonymizer to ≥2.2.362
• ✅ Fixed critical transformers vulnerability (GHSA-phhr-52qp-3mj4)
• ✅ Fixed Anonymize scanner to properly respect language parameter (#337)
• ✅ Fixed PromptInjection tokenizer initialization bug (#331)
• ✅ Fixed MaliciousURLs missing top_k parameter TypeError (#318)

New Features & Enhancements:

• ✅ Python 3.13+ Support: Full compatibility with Python 3.13 and 3.14
• ✅ ATRScanner: New Agent Threat Rule scanner with 10+ built-in detection patterns
• ✅ ThresholdMixin: Dynamically adjust scanner sensitivity without model reloading
• ✅ Model.from_local(): Factory method for locally-downloaded HuggingFace models
• ✅ Improved dependency management for better compatibility

Get Involved

• Give us a ⭐️ github star ⭐️ on the top of this page to support what we're doing, it means a lot for open source projects!
• Read our docs for more info about how to use and customize LLM Guard Rails, and for step-by-step tutorials.
• Post a Github Issue to submit a bug report, feature request, or suggest an improvement.
• To contribute to the package, check out our contribution guidelines, and open a PR.

Support & Contact

We're eager to provide personalized assistance and discuss feature requests or improvements.

• Send Email ✉️ - Open to feature requests, improvements, and collaboration
• GitHub Issues - Report bugs or request features
• GitHub Discussions - Ask questions and share ideas

Acknowledgments

This project is a maintained fork of the excellent LLM Guard by Protect AI. We maintain this fork to provide:

• Active Security Updates: Prompt patches for CVEs and security vulnerabilities
• Regular Bug Fixes: Community-reported issues resolved quickly
• New Features: Enhancement requests implemented with community feedback
• Continuous Support: Regular releases and dependency updates

The original LLM Guard project by Protect AI deserves recognition for building this solid foundation.

License

This project is licensed under the MIT License - see the LICENSE file for details.