Protego field desk

Secret Leak Scanner

Check pasted code, config, logs or .env-style text for likely leaked credentials. Free, local and no login required.

Quick answer

Paste code, configuration or logs to check for recognizable credential patterns without uploading the text. This free secret leak scanner runs in your browser and masks findings with line numbers. It cannot confirm that a credential is active. Revoke exposed credentials first, then remove copies and review repository history.

Text stays in this browser tab. No upload, persistence or credential verification. Maximum 500 KB in UTF-8 bytes. Results show only the first 4 and last 2 characters.

0 / 500,000 bytes

Rules, sources and limitations

Patterns checked 2026-11-16. Pinned Gitleaks default rules inform this curated subset. It does not include all Gitleaks rules or allowlists.

The generic assignment heuristic uses Shannon entropy ≥ 3.5 bits per character on values of 16 to 256 characters. It can flag random identifiers and miss short, encoded or unfamiliar credentials. AWS IDs alone are not sufficient to authenticate.

Revoke or rotate real exposed secrets first, then coordinate history cleanup. This is not a replacement for Gitleaks or TruffleHog in CI.

Frequently asked questions

Does the secret scanner upload or save my text?

The scan runs locally in your browser. Pasted text and matched values are not sent by the tool or saved to browser storage. Session replay is excluded on this page. Usage analytics contain only the tool name, never the pasted text.

Which credential patterns can it detect?

The curated rules cover AWS access key IDs, GitHub and GitLab tokens, selected Slack tokens and webhooks, Stripe secret and restricted keys, Google API keys, documented OpenAI-style formats, private key PEM blocks and JWT patterns. A separate entropy heuristic checks credential assignments. The rule references are linked below.

How does the entropy heuristic work?

Assignments to names containing password, secret, token or api_key are checked when the value has 16 to 256 supported characters. Values with Shannon entropy of at least 3.5 bits per character are flagged. This is a local heuristic, not a provider format or a measure of password strength. Random identifiers can match.

What should I do after finding a leaked secret?

Revoke or rotate a real exposed credential first, update authorized applications, and review activity at its issuer. Then purge exposed copies and, where needed, coordinate repository history cleanup. Deleting a line from the latest commit does not revoke a credential.

Can this replace Gitleaks or TruffleHog in CI?

No. This pattern-based paste check can produce false positives and false negatives. It does not scan a repository, inspect Git history, decode every encoding or verify credentials. Use a maintained repository scanner in CI. Input is limited to 500 KB in UTF-8 bytes and results stop at 200 findings.

Related tools and guides