PixelOkay
Local-only processing
Tools
DEVELOPER TOOLS SECURITY

Debug Evidence Sanitizer

Remove credentials, private data and sensitive request headers from logs, stack traces and issue reports before you paste them into tickets, chats or AI assistants.

🔒 Runs in your browser◎ No account or upload↗ Sanitized exports
SANITIZER WORKSPACE

Make your evidence share-ready

Import text files or paste a report. Review the result before sharing.

Privacy by designYour source stays in this tab. No analytics, external scripts, server processing or automatic storage.
01 / SOURCE

Original evidence Sensitive

0 characters
Original content is never included in exports.
02 / OUTPUT

Sanitized evidence Review before sharing

Not scanned
Run the sanitizer to generate a clean draft. You can manually refine it before export.
CONFIGURATION

Redaction preferences

Credentials always on
Optional checks can produce false positives.

Custom values are case-sensitive and stay in this tab. Names, addresses and unknown secrets may require manual terms.

YOUR SCAN

Redaction report

—Redactions
—Categories
—Lines
Ready for source text. Scan manually when you’re set.

Findings by category

Select a finding category to jump to its first redaction; select again to review the next. Only up to 3,000 findings are listed. Manual output edits invalidate the automatic finding positions.

Categories and line references will appear here. No secret values are shown in this list.
Manual review is essential. Pattern matching can miss secrets hidden in free text, encoded data, unusual formats, or images. Do not treat a zero-match result as proof of safety.
HOW IT WORKS

Designed for debugging, not data collection.

01

Add your evidence

Paste a stack trace, import HTTP logs, HAR exports or plain-text issue notes. Multiple text files can be combined without preserving filenames.

02

Detect and redact

Local rules flag credentials, tokens, private keys, request headers, email addresses and optional personal data. Configure replacements or add exact terms.

03

Review, then share

Check the entire sanitized draft, see which lines were changed, then copy or export only the sanitized text and metadata.

IMPORTANT LIMITATIONS

A safer starting point. Not a guarantee.

This is a deterministic, rules-based sanitizer built for AI-ready debugging evidence; it does not send your data to an AI model or promise comprehensive PII detection.

Text inside images, encrypted values, base64-wrapped secrets, unfamiliar credentials, free-form names, addresses, proprietary identifiers and sensitive business context may remain. Rotate any credentials that have already been exposed.

Nothing is stored by this tool, but your browser, OS clipboard and exported files can retain copied content. Review results before posting to public issues or third-party assistants.

Using Debug Evidence Sanitizer

Find recognizable credentials and optional personal data in local debugging text, then review and export a redacted copy. Treat findings as prompts for inspection rather than proof that the text is safe.

  1. Add up to ten UTF-8 text files or paste debugging evidence, then choose the credential and personal-data patterns to check.
  2. Review each finding, navigate to its context and adjust custom values or replacement style; inspect the full sanitized output, including areas with no findings.
  3. Export sanitized TXT, a JSON finding report or issue Markdown after checking that the replacements preserve useful debugging context.

Example: Paste a short request log, add a project-specific token to the custom-value list, review each proposed replacement in context and export the sanitized text for a ticket.

Limits and privacy: Detection is incomplete: zero findings do not prove that evidence is safe to share. Review the full output and surrounding context. Inputs are limited to ten UTF-8 files and 5,000,000 combined bytes. Custom values and patterns may miss secrets or sensitive context. This release uses local rule-based processing without remote AI; it does not automatically persist imported evidence.

Frequently asked questions

Does an empty findings list mean my logs are safe to share?

No. The rules recognize patterns and values you provide, but can miss secrets or personal information. Review the entire sanitized output and consider the surrounding context before sharing.

Are my debugging files uploaded or saved automatically?

The redaction worker processes the text locally and this release does not use remote AI or automatically persist imported evidence. Exports you create can contain original or sanitized text, so inspect them before sharing.

Related tools