Tools / Secret & API key scanner
Secret & API key scanner
Paste code, a config file or a JavaScript bundle — or open files — and find leaked API keys, tokens, private keys and credentials before they ship. The scan runs entirely in your browser.
Each file is capped at 5 MB and read in your browser. Minified bundles are fine.
Findings
| Severity | Type | Location | Match (masked) | Confidence |
|---|
What this scans for
It matches the formats of well-known credentials — AWS, Google/GCP, Firebase, Stripe, GitHub, GitLab, Slack,
Twilio, SendGrid, OpenAI, Anthropic, npm, PyPI and more — plus private-key blocks, JSON Web Tokens, database
connection strings with embedded passwords, credentials in URLs, and generic key = "…" assignments.
A final pass flags high-entropy strings that look like secrets even when the provider isn't recognised.
How to read the results
Provider-format matches (e.g. an AKIA… AWS key) are high confidence. Generic assignments and
high-entropy strings are lower confidence and include false positives — a long hash, a content-security nonce or
a cache-busting token can look like a secret. Each finding shows its confidence so you can triage. The match is
always masked; the line number points you to the original.
Limits worth knowing
This tool only sees what you give it. It cannot find a secret in a file you didn't paste, and finding none is not proof your code is clean. If a real secret is flagged, treat it as compromised the moment it reached a shared repo or a shipped bundle: rotate it first, then remove it from the source and move it to an environment variable or a secrets manager. Rotating matters more than deleting — once committed or shipped, assume it's already been harvested.