The Zinc Philosophy: Sacrificial Security for AI Agents
As autonomous agents execute tools, modify codebases, and interact with the open internet, zn acts as the sacrificial boundary between untrusted inputs and agent runtime.
Why Zinc (Element 30)?
AI agents are becoming the active operating layer between humans and critical systems. They read unstructured documents, browse live web pages, execute bash commands, and trigger MCP tools on our behalf. But this autonomy introduces severe systemic vulnerability.
In metallurgy, galvanizing steel with a thin zinc layer provides sacrificial anode protection: zinc oxidizes first, taking the electrochemical damage so the underlying structural steel remains completely untouched.
We built zn to bring this exact physical defense to software agents. Sitting as a zero-latency, local-first proxy, zn absorbs direct prompt injections, poisoned tool outputs, and credential harvesting attempts before hostile payloads ever reach your LLM context or shell.
Our Engineering Principles
Zero External Dependencies & Local-First
Security tooling should never expand your attack surface. zn-gate runs 100% offline with zero external npm dependencies, weighing under 15 kB and deciding in sub-millisecond speeds (< 0.1 ms) directly on your local CPU.
Full-Lifecycle Bidirectional Coverage
Protection must extend beyond the user prompt. zn inspects incoming user inputs (analyze_prompt), pre-flight tool call arguments (check_tool_call), and post-execution third-party outputs (check_tool_result) against Indirect Prompt Injection.
Deterministic Speed + Deep Neural Camera
Immediate regex pre-normalization (homoglyphs, zero-width chars, inline comments, Base64 decoding) catches active evasions instantaneously, fused with the certified Galvanize-60M dynamic INT8 ONNX engine (1.00% tool FPR, 91.60% Deepset OOD recall) for complex zero-day vectors.
Open Science & Reproducibility
No security through obscurity. We openly publish our 23.7k benchmark dataset on Hugging Face, our gate package on npm, and our code on GitHub. Every performance claim is independently verifiable.
Supported by Global Cloud & AI Programs
zn builds on enterprise-grade sovereign identity, accelerated inference, and GPU compute provided by top global cloud & AI accelerators.
AWS for Startups (Activate)
zn is built natively on AWS cloud primitives — CloudFront global edge CDN, S3, Route53, DynamoDB, and Serverless Lambda with Secrets Manager — supported by AWS Activate credits and technical architecture guidance.
Universal MCP Architecture
Designed from day one for the Model Context Protocol (MCP). Out-of-the-box native compatibility with Cursor, Claude Code, Antigravity, OpenCode, Codex, Hermes Agent, and custom LLM workflows.
NVIDIA Developer
Optimized neural inference engines, int8 quantization, and TensorRT acceleration for ultra-low latency prompt classification.
Nebius AI Builder Program
Cloud GPU infrastructure powering continuous model training, adversarial evaluation, and certified Galvanize-60M engine evaluation.
Auth0 for Startups
Sovereign authentication, secure token lifecycles, and cryptographic authorization standards for agent API gateways.
Multi-region deployment across AWS, with support from the AWS Activate program.
What we protect
Consider an operations agent reading a support ticket or web scrape that contains an instruction to run rm -rf /. zn gives the agent a sacrificial boundary and an evidence trail instead of blindly assimilating external strings as system commands.
| Approach | Outcome |
|---|---|
| zn security gate | Blocks injected tool calls, sanitizes third-party data, and preserves audit trails |
| Traditional WAF | Inspects transport and IP patterns, but blind to semantic LLM hijacking |
| No protection | Untrusted web/user strings reach agent shell and tools directly |
Security & Responsible Disclosure (RFC 9116)
We practice coordinated vulnerability disclosure. If you discover a potential security flaw, guardrail bypass, or evasion vulnerability in zn-gate or our cloud infrastructure, please report it directly to our security team. We provide safe harbor for researchers acting in good faith.
Official Directory & Contact
Platform integrations, partnership inquiries, and product questions.
Vulnerability reports, penetration testing findings, and security audits.
Direct contact with the zn engineering team for architectural discussions.
Ready to protect your agents?
Start free with a 7-day trial. No credit card required.