Start with Identity
Control · 5 briefs

Attacks on credential storage

Techniques that target where secrets rest rather than how they are used: infostealer malware, secrets left in source repositories and CI logs, and key extraction from the vaults meant to protect them.

Why this control fails

Credentials that sit still are easier to steal than credentials in motion. Infostealer malware harvests browser-saved passwords, session cookies, and API keys from a single compromised endpoint in one pass. Source repositories and CI pipeline logs leak long-lived static secrets routinely, because a secret committed once and later removed from the file usually still exists in git history. Secrets managers themselves are not exempt: an authenticator bug or an over-permissioned policy engine on a vault turns "we store secrets securely" into "we handed out every secret the vault held," which is a worse outcome than the sprawl the vault was bought to fix.

What security people should do

  • Scan repositories and CI logs for committed secrets continuously, and treat a found secret as compromised, not just as a finding to close.
  • Replace long-lived static API keys with short-lived, automatically rotated credentials wherever the platform supports it.
  • Patch secrets manager software on the same cadence as any other identity provider, and restrict who can edit authenticator configuration or policy engine templates.
  • Rotate every secret a vault or credential store held if its authenticator was reachable while a known vulnerability was unpatched.

Techniques in this category

Seeing this attack in production and have a detection idea we should add? Email [email protected] or volunteer as a CVE Analyst.