Open-source AI agent firewall for MCP security and agent egress. Scans mediated HTTP, MCP, A2A, and WebSocket traffic for exfiltration, SSRF, and prompt injection, and emits mediator-signed action receipts: verifiable audit evidence from outside the agent.

The Open Source Firewall for LLMs. A self-hosted gateway to secure and control AI applications with powerful guardrails.

Open-source security gateway for LLM APIs — prompt injection detection, PII redaction, dangerous response sanitization, and audit logging. OpenAI/Claude compatible, MCP & Agent SKILL support. Drop-in proxy for AI coding agents (Cursor, Claude Code, Codex).

Ship AI agents with guardrails — not prayers. Self-hosted runtime protection for LLMs and tool-calling agents: block prompt injection, enforce tool permissions, redact sensitive data, and control what agents are allowed to do.

Self-hosted LLM security proxy. PII redaction, prompt injection defense, KVKK/GDPR/PCI-DSS compliance. Sub-millisecond latency.

LLM Security Platform.

Deterministic AI defense framework: immutable input filtering, n-model cryptographic hash consensus, and self-learning adaptive rules. Zero dependencies. Hardware-sealed. Patent Pending.

Open-source AI security firewall. 81 engines for PII detection, prompt injection defense, MCP security, and egress classification. Local-first. Zero cloud dependency.

pre-alpha: An inline LLM firewall with a sub-10 ms p99 latency target — built in layers across five documented phases. Sits between your app and any LLM endpoint to classify, redact, or block threats in real time, then continuously retrains itself when drift is detected.

Prompt-injection firewall for LLM applications — 33 input detectors, 9 output scanners, federated ed25519-signed threat-intel feed. Apache 2.0, 1040 tests, F1 96.0% with 0% false positives. Docker, GitHub Action, LangChain/LlamaIndex/CrewAI integrations.

Pure-Rust prompt-injection detector with 1.5MB embedded MLP classifier. 98.40% accuracy, p50 14ms CPU inference, bindings for Python/JS/Go. Apache-2.0/MIT alternative to Rebuff (archived) and Lakera Guard.

AegisMCP Enterprise is a dual-layer security gateway and cost firewall designed to sit between AI models (like Claude and Cursor) and execution engines.

🛡️ AI-Guard: The Open-Source Security Gateway & Audit Platform for AI Agents. Block injections, control permissions, and track LLM costs.

LLM Security Platform Docs

Portable runtime policy and audit layer for AI agents - HTTP/HTTPS proxy enforcing egress policies, inspecting content, materializing secrets, and recording every decision.

Enterprise-grade prompt injection detection and AI firewall for LLM applications

Open-source LLM firewall — drop-in OpenAI-compatible proxy with layered policy engine, jailbreak detection, audit logs, and a dashboard. Apache 2.0.

Input/output safety firewall for LLM apps: PII redaction (Luhn-validated), prompt-injection/jailbreak detection, secrets + toxicity filters, with a labeled precision/recall benchmark. FastAPI + Claude.

🛡️ Detect and block prompt injection attacks in LLM apps using pattern detectors, ML, and community-driven feedback to improve security.

Semantic Prompt Injection Firewall for Large Language Models using Sentence Transformers and XGBoost