Security proxy for AI agents. Scans every message for prompt injection, PII, and secrets. Defense-in-depth: Go proxy + iptables firewall + eBPF kernel monitor. YAML policy engine, audit logging, 5 AI agents with RAG knowledge bases.

The Anti-Virus for AI Artifacts & RAG Firewall. A static analysis tool scanning Models and Notebooks for RCE, Datasets and RAG docs for Data Poisoning, PII, and Prompt Injections. Secure your AI Supply Chain.

Client-side retrieval firewall for RAG systems — blocks prompt injection and secret leaks, re-ranks stale or untrusted content, and keeps all data inside your environment.

Open-source LLM Security Firewall with fine-tuned threat detection, RAG-based threat intelligence, and proxy architecture. Built for secure deployment of Large Language Models using zero-cost open-source tools.

SAFE-RAG++ is a research-grade pipeline to diagnose, quantify, and reduce hallucinations in retrieval-augmented generation (RAG). It blends causal interventions with a multi-signal firewall (NLI entailment · self-consistency · token-level uncertainty).

ZeroShield AI Mesh Firewall is a centralized AI security gateway for governing LLM and RAG traffic with inline prompt injection defense, vector database isolation, multi model routing control, and compliance grade observability.

No description available

Adversarial-Grade Defense Layer for Retrieval-Augmented Generation Systems.

The AI Firewall is a production-grade security microservice that acts as a gatekeeper, ensuring that your LLM remains an asset rather than a liability.

A consistency-based firewall for high-stakes Retrieval Augmented Generation (RAG). Queries the model multiple times and incinerates the output if entropy is high (divergent answers), preferring silence over hallucination.

OPA-style policy engine for secure RAG — sensitivity labels, retrieval rules, chunk trust levels, prompt injection defense, and full decision evidence

No description available

No description available

No description available

Phase 2: Civitas RAG (UPSC knowledge engine) + LLM Firewall (PII redaction, injection detection, cost tracking)

No description available

Code for paper "CONTROLNET: A Firewall for RAG-based LLM System"

🛡️ Enterprise AI semantic firewall — intercepts LLM prompts, auto-masks PII, enforces corporate data policies via RAG, and routes sensitive requests to private Azure deployments.

Hallucination Firewall is an MLOps system that detects and controls LLM hallucinations using a RAG pipeline, similarity scoring, and Logistic Regression. It classifies responses (SAFE/WARNING/HALLUCINATION/OUT_OF_SCOPE), applies a control layer, and supports web fallback for out-of-scope queries.