Security Tools for AI Agents

Install. Enable. Done.

No MCP server setup. No client configuration. Security tools appear automatically when you enable the proxy. Any LLM routing through gets the tools.

Coming Soon
SDK and PTI integration launching soon. Join the waitlist for updates.

Two Ways to Integrate

PTI (Proxy Tool Injection)

Zero Code Integration — Coming Soon

Enable the proxy and tools automatically appear in any LLM routing through it. Works with Claude Code, Cursor, custom agents - anything using OpenAI or Anthropic APIs.

Coming Soon
  • No SDK import required
  • No MCP server setup
  • No client configuration
  • 40+ security tools injected automatically
  • Supply chain protection included

Python SDK

Direct Integration — Coming Soon

Import the SDK directly in your Python code for programmatic access to all security tools.

Coming Soon
  • Type hints included
  • Full scanning + Shield
  • Auto-escalation logic
  • Best for Python projects

How PTI Works

Unlike MCP which requires separate servers and client configuration, PTI injects tools directly into API traffic.

1

Install

Install the package

Coming Soon
2

Enable

Start the proxy

Coming Soon
3

Use

Tools appear automatically

LLM calls: scan, shield_scan...

Discover All 40+ Tools

40+ security tools will be available for content scanning, device hardening, and supply chain protection:

Coming Soon
Join the waitlist to be notified when the SDK launches

Quick Reference (for agents)

Concise SDK/PTI guide that agents can fetch directly:

GET /v1/skill.md

Full HTTP API

Complete API documentation for direct HTTP integrations:

GET /v1/api.md

Get Notified at Launch

Be the first to know when the SDK is available.

Join the Waitlist Get launch updates