securevector-ai-monitor 2.0.0

Real-time AI threat monitoring. Protect your apps from prompt injection, leaks, and attacks in just a few lines of code.

SecureVector

Runtime Firewall for AI Agents & Bots

Block prompt injection, jailbreaks, and data leaks before they reach your AI.

Website · Docs · Demo · Getting Started · Use Cases · API · Discord

How It Works

SecureVector sits between your AI agent and the LLM provider, scanning every request and response for security threats. Runs entirely on your machine — nothing leaves your infrastructure.

pip install securevector-ai-monitor[app]
securevector-app --web

Or download: Windows · macOS · Linux

Open-source. 100% local. No API keys. No cloud. No data sharing.

Highlights

• ☑ 100% Local — No data transmitted externally. Complete privacy.
• ☑ Agents Protected — LangChain, LangGraph, CrewAI, n8n, OpenClaw, and any OpenAI-compatible app.
• ☑ Input Scanning — Block prompt injection, jailbreaks, and manipulation before they reach the LLM.
• ☑ Output Scanning — Detect credential leaks, PII exposure, and system prompt disclosure.
• ☑ 18+ Providers — OpenAI, Anthropic, Gemini, Ollama, Groq, Azure, and more.
• ☑ One Command — securevector-app --web and follow the UI to start protecting.

Install

Runtime: Python 3.9+ (MCP requires 3.10+)

Install	Use Case	Size
pip install securevector-ai-monitor[app]	Local app — dashboard, LLM proxy, self-hosted	~60MB
pip install securevector-ai-monitor	Cloud SDK — lightweight, uses cloud API	~6MB
pip install securevector-ai-monitor[mcp]	MCP server — Claude Desktop, Cursor	~20MB

# Local users (self-hosted, OpenClaw proxy)
pip install securevector-ai-monitor[app]
securevector-app

# Cloud users (API integration)
pip install securevector-ai-monitor

Binary installers: Windows · macOS · Linux · All Releases

Quick Start

Step 1: Start SecureVector app

securevector-app --web

Step 2: Go to Integrations in the UI, choose your agent framework and LLM provider, then click Start Proxy.

Step 3: Point your app to the proxy (shown in the UI).

That's it! Every request is scanned for prompt injection. Every response is scanned for data leaks.

Supported providers: openai anthropic gemini ollama groq openrouter deepseek mistral xai azure together fireworks perplexity cohere cerebras lmstudio litellm

Agent Integrations

Agent/Framework	Integration
LangChain	LLM Proxy or SDK Callback
LangGraph	LLM Proxy or Security Node
CrewAI	LLM Proxy or SDK Callback
Ollama / Open WebUI	LLM Proxy — see Integrations in UI
OpenClaw / ClaudBot	LLM Proxy — see Integrations in UI
n8n	Community Node
Claude Desktop	MCP Server Guide
Any OpenAI-compatible app	LLM Proxy — set OPENAI_BASE_URL to proxy
Any HTTP Client	POST http://localhost:8741/analyze with {"text": "..."}

What It Detects

Input Threats (User → LLM)	Output Threats (LLM → User)
Prompt injection	Credential leakage (API keys, tokens)
Jailbreak attempts	System prompt exposure
Data exfiltration requests	PII disclosure (SSN, credit cards)
Social engineering	Jailbreak success indicators
SQL injection patterns	Encoded malicious content

Full coverage: OWASP LLM Top 10

Screenshots

Dashboard — stats, risk distribution, recent threats	Threat Analytics — blocked, redacted, logged
Integrations — LangChain, Ollama, OpenClaw, and more	LLM Proxy — provider configuration

Documentation

• Installation Guide — Binary installers, pip, service setup
• Use Cases & Examples — LangChain, LangGraph, CrewAI, n8n, FastAPI
• MCP Server Guide — Claude Desktop, Cursor integration
• API Reference — REST API endpoints
• Security Policy — Vulnerability disclosure

Editions

Open Source	Professional/Enterprise
Apache 2.0 license	Expert-curated rule library
Community detection rules	Multi-stage ML threat analysis
Custom YAML rules	Real-time cloud dashboard
100% local, zero data sharing	Team collaboration
Desktop app + local API	Priority support & SLAs

Try Free · Pricing · Enterprise

Contributing

git clone https://github.com/Secure-Vector/securevector-ai-threat-monitor.git
cd securevector-ai-threat-monitor
pip install -e ".[dev]"
pytest tests/ -v

Contributing Guidelines · Code of Conduct

License

Apache License 2.0 — see LICENSE.

SecureVector is a trademark of SecureVector. See NOTICE.

---

Get Started · Documentation · GitHub Issues · [e-mail]