resk-llm is a free open-source Python library that wraps OpenAI-compatible APIs with a security layer, blocking prompt injection, PII leaks, and malicious URLs while managing context.

Are the RESK security libraries really open source?

Yes. All RESK libraries including resk-llm, resk-llm-ts, resk-mcp, reskSecure, resk-logits, resk-caching, and ReskPoints are published on GitHub and distributed for free on PyPI and npm.

How do I install resk-llm?

Install resk-llm with pip: pip install resk-llm. For JavaScript, use npm install resk-llm-ts. For caching and vector search, use npm install resk-caching.

8 Open Source Libraries to Secure Your LLM Applications

Building AI applications is easy. Securing them is hard. That is why RESK Security publishes its entire security stack as free open-source libraries on GitHub, available on PyPI and npm. Whether you write Python, TypeScript, or Bun, there is a RESK library for your LLM stack. Here is the complete ecosystem.

Python Libraries (PyPI)

resk-llm - Python LLM Security

pip install resk-llm - The flagship library. Wraps OpenAI, Anthropic, and Cohere APIs with a protective layer that blocks prompt injection, PII leaks, and malicious URLs, manages context, and enforces rate limits.

GitHub: Resk-Security/Resk-LLM | PyPI | Docs

reskSecure - Bitmask LLM Firewall

A bitmask-based LLM security firewall with policy-driven logits filtering using GPU-accelerated Aho-Corasick pattern matching. Blocks dangerous content during generation at the token level.

GitHub: Resk-Security/reskSecure

resk-logits - Shadow Ban Processor

A logits processor that implements a "shadow ban" system to filter dangerous content during text generation. Silently prevents harmful completions without breaking the conversation flow.

GitHub: Resk-Security/resk-logits

resk-mcp - MCP Security Layer

pip install mcp-resk - Adds a robust security and management layer over the official Model Context Protocol SDK. Secures tool calls, validates protocol interactions, and monitors MCP activity.

GitHub: Resk-Security/resk-mcp | PyPI

ReskPoints - AI Agent Logger

pip install reskpoints - Track every action your AI agents take: function name, parameters, token probabilities, and results. Ship to Datadog, Prometheus, OpenTelemetry, webhooks, or files with async batch processing.

GitHub: Resk-Security/ReskPoints | PyPI

TypeScript & Bun Libraries (npm)

resk-llm-ts - TypeScript LLM Security

npm install resk-llm-ts - A comprehensive security toolkit for JavaScript and TypeScript applications using LLMs. Wraps OpenAI-compatible APIs with threat detection, rate limiting, and caching.

GitHub: Resk-Security/resk-llm-ts | NPM

resk-caching - Bun Caching & Vector Search

npm install resk-caching - A Bun-based backend library and server for secure caching, embeddings orchestration, and vector search. Built for AI chatbot backends with Redis integration.

GitHub: Resk-Security/resk-caching | NPM

Mobile & Research

ReskCrypt - Encrypted AI Conversations

End-to-end encryption for AI conversations with policy-based, revocable access control. Protect sensitive data shared with chatbots and LLM applications.

GitHub: Resk-Security/ReskCrypt

Why Contribute to RESK?

Free and MIT-style: All libraries are open source and free to use.
Battle-tested: Used in production AI security deployments.
Active development: New releases and features shipped regularly.
Community-driven: Issues, PRs, and discussions are welcome.

Get Started

Browse all repositories at github.com/Resk-Security, star your favorites, and open an issue if you find a bug. Every contribution helps make AI safer for everyone.