SecureMidSecurityAuto-graded

Redact live credentials from exported support logs

Catching every disguise a credential wears in a support log, while leaving ordinary text untouched, is the bar.redactor.py is the scrubber behind the support bundle: when a customer hits a bug we ask them to paste the tool's output into a ticket, and redact(text) is supposed to strip every secret before it leaves their machine. That output routinely contains live credentials, and the shipped version leaks them. Replace every secret with the redaction marker while leaving all the surrounding text — the part that makes the bundle useful — completely intact.

30 min

python

Start lab

You're done when

Secrets are fully removed, with no fragment of the original value left behind.
Cloud access keys are redacted.
Bearer tokens and JWTs are redacted.
Credentials appearing as key-value pairs are redacted.
Long hex and base64 secrets are redacted.
Card numbers are redacted.
Values that are not secrets are preserved, and benign lines pass through unchanged.

What it measures

Application securityRecognising the full range of credential shapes that appear in real diagnostic output.

Security gap analysisCovering every secret format rather than the obvious ones, since a single missed pattern leaks the credential entirely.

Redaction without destructionKeeping the non-secret text intact, because a bundle scrubbed to uselessness is a fix nobody ships.

Your environment

Terminal + editorA full shell in the browser

Claude CodePre-authenticated, no key setup

Isolated microVMYours alone, wiped afterwards

Start this lab

30 min

Start labYou'll sign in to Anthropos to start. The lab opens in a private cloud workspace with Claude Code ready to use.

AI Labs is part of Anthropos Workforce. Not on your plan yet? Book a demo