Veild - AI Code Privacy

About

Code with AI without leaking company secrets.

Veild is a zero-knowledge security layer that sanitizes your code before it leaves your clipboard. It allows developers to safely use AI assistants (ChatGPT, Claude, Perplexity and Gemini) with proprietary code by removing sensitive data client-side.


The Core Loop: Redaction & Rehydration Veild solves the "broken code" problem common with other obfuscators.

• On Paste (Redaction): When you paste code into an AI chat, Veild intercepts the event and uses veild engine analysis to replace API keys, PII, and credentials with typed placeholders (e.g., [VEILD_REDACTION_1]). The AI receives safe, structural code.
• On Copy (Rehydration): When you copy the AI's response back to your IDE, Veild detects the placeholders and instantly restores your original values locally. Your code remains functional with zero manual cleanup.


Enterprise-Grade Privacy
• True Local Execution: The entire engine runs locally within your browser.
• Zero Network Requests: Veild has no backend. No analytics, no logging, and no data egress. You can verify this via the Network tab in DevTools.


Capabilities
• Invisible Workflow: Veild operates invisibly. It intercepts your Cmd+V (Paste) and Cmd+C (Copy) events directly in the browser to sanitize and restore code instantly. There are no popups to approve, no UI overlays to navigate, and zero context switching. Your workflow remains exactly the same, just that now is secure.
• Broad Language Support: Native support for Plain Text, Python, TypeScript, Java...


Supported Platforms
• OpenAI (ChatGPT)
• Anthropic (Claude)
• Perplexity
• Google Gemini



Why Veild? Public AI models train on user data. Pasting internal IP addresses, sk_live keys, or customer emails puts your organization at risk. Veild eliminates this vector completely, allowing you to leverage AI productivity without compromising security posture.

Veild is free and open source for individual developers.