⟲

Code Transformer & LLM Sanitizer

transform.ignlab.net

100% Local Processing • Zero Data Transmitted • Safe for Proprietary Code
Mode
AI Safeguard {{ redactions.length }} redacted
Indent
No history yet — copy or download a result to save it here.
Source {{ activeMode.from }} detected: {{ detectedLabel }}
Drop file to load…
{{ inputGutter }}
{{ inputError }} {{ stats.in.lines }} lines · {{ stats.in.chars }} chars · {{ stats.in.bytes }}
Output {{ activeMode.to }} 
{{ output }}{{ outputError ? '' : 'Transformed output appears here…' }}
⚠ {{ outputError || copyError }} {{ stats.out.lines }} lines · {{ stats.out.chars }} chars · {{ stats.out.bytes }}
🛡 Sanitizer redacted {{ redactions.length }} sensitive item{{ redactions.length===1?'':'s' }} before transforming
{{ r.name }} ×{{ r.count }}

AI Safeguard — sanitize code before pasting it into an AI assistant

AI Safeguard is a one-toggle sanitizer that scans any pasted text for private key blocks, JWTs, AWS/GCP/Azure/Stripe/GitHub/OpenAI/Slack/npm tokens, bearer tokens, credential assignments (password=, api_key:), credit card numbers, emails and IP addresses — and redacts them before you copy your text out. Everything runs locally in your browser.

By assistant

  1. Pasting code into ChatGPT?
  2. Pasting code into Claude?
  3. Pasting code into Gemini?
  4. Pasting code into GitHub Copilot Chat?

Frequently asked questions

What gets redacted?

Private key blocks, JWTs, AWS/GCP/Azure/Stripe/GitHub/OpenAI/Slack/npm tokens, bearer tokens, credential assignments (password=, api_key:), credit card numbers, emails and IPv4 addresses.

Does it also convert my code’s format?

You can combine it with any conversion mode — sanitize first, then the cleaned text flows into whichever converter is selected — or use it standalone with any format.

Is anything uploaded?

No, sanitization runs entirely client-side with regular expressions; nothing you paste is ever sent anywhere.

Want a full audit first? DataExtract's API key extractor lists every exposed credential it finds, so you can see exactly what AI Safeguard is about to redact.