guardrails-ai 0.10.1 malicious release
With an employee's GitHub token, the attacker reached the project's publishing credentials and released a malicious guardrails-ai 0.10.1 to PyPI. It carried the self-spreading credential stealer of the May 2026 Mini Shai-Hulud wave, and PyPI quarantined it within about two hours.
- First seen
- 2026-05-12
- Disclosed
- 2026-05-12
- Affected packages
- 1
- Attributed to
- TeamPCP, according to OSV (Google Open Source Security) and StepSecurity
What to do
Pin to 0.10.0 or a later release (0.10.2 or 0.11.0). If 0.10.1 was installed, treat the host as compromised: rotate GitHub, cloud and registry tokens and API keys, and look in GitHub for workflows, repositories or commits you did not make.
Check a lockfile for these releases →Affected releases
| Package | Registry | Versions |
|---|---|---|
| guardrails-ai | PyPI | 0.10.1 |
On ShieldMap
Sources
- https://github.com/guardrails-ai/guardrails/blob/main/SECURITY_ADVISORY.md · Guardrails AI (maintainers), 2026-05-12
- https://github.com/guardrails-ai/guardrails/security/advisories/GHSA-xmpw-2vmm-p4p6 · Guardrails AI / GitHub Advisory Database, 2026-05-19
- https://osv.dev/vulnerability/MAL-2026-3607 · OpenSSF Malicious Packages (via OSV), 2026-05-12
- https://www.stepsecurity.io/blog/mini-shai-hulud-is-back-a-self-spreading-supply-chain-attack-hits-the-npm-ecosystem · StepSecurity, 2026-05-11
Written up from the sources above and last checked on 2026-11-18. ShieldMap describes what was affected and what to do; it does not publish the malicious code or how it worked. Report a correction to [email protected].