Skip to main content
Mallory
Critical

Malicious PyPI package in Guardrails AI 0.10.1

IdentifiersCVE-2026-45758CWE-506· Embedded Malicious Code

CVE-2026-45758 concerns a malicious publication of the Python package guardrails-ai version 0.10.1 to PyPI. On 2026-05-11 at approximately 18:00 Pacific, an attacker published the malicious 0.10.1 release. Any user who installed guardrails-ai==0.10.1 from PyPI on that date may have been exposed. The package was identified as malicious by security researchers within about two hours, after which PyPI quarantined the repository. The available information indicates this is a software supply-chain compromise affecting a specific package version rather than a flaw in the legitimate Guardrails AI codebase. The malicious package could access credentials available on the installing machine and may have targeted developer environments and associated GitHub resources.

Share:
For your environment

Are you exposed to this one?

Mallory correlates every CVE against your assets, your vendors, and active adversary campaigns. Know which vulnerabilities matter for you, not just which ones are loud.

ANALYST BRIEF

Impact, mitigation & remediation

What it means. What to do now. Patch path, mitigations, and the assume-compromise checklist.

Impact

What an attacker gets, and what they’ve been doing with it.

Successful exposure to the malicious package may have resulted in compromise of credentials accessible from the affected host, including GitHub personal access tokens, cloud provider keys, package registry tokens, and other API keys. The advisory also indicates potential follow-on abuse of GitHub accounts, including unauthorized workflows or repositories. Based on Guardrails AI telemetry and review of system and access logs, the maintainers observed no requests to Guardrails AI infrastructure originating from the malicious 0.10.1 version and found no evidence of user data exfiltration through their systems; however, compromise of local secrets and downstream third-party services remains the primary impact.

Mitigation

If you can’t patch tonight, do this now.

Avoid installing guardrails-ai==0.10.1 from PyPI. If installation status is uncertain, verify package versions across developer endpoints, build systems, and CI/CD pipelines. Restrict and minimize credentials present on developer machines and runners, use short-lived tokens where possible, and monitor GitHub and cloud accounts for anomalous activity. Organizations should validate package provenance, pin to trusted versions, and use repository and dependency monitoring controls to detect malicious or unexpected package releases.

Remediation

Patch, then assume compromise.

Remove the malicious package version from affected environments and replace it with a known-good release. Users should upgrade to guardrails-ai version 0.10.2 or downgrade to version 0.10.0, both of which are identified as unaffected. Any system on which guardrails-ai==0.10.1 was installed should be treated as potentially exposed: rotate all credentials that may have been accessible from that machine, including GitHub PATs, cloud provider credentials, package registry tokens, and API keys. Review GitHub accounts and organizations for unauthorized workflows, repositories, or other suspicious changes, and perform broader incident-response validation on affected developer workstations or CI environments as appropriate.
PUBLIC EXPLOITS

Exploits

No public exploits tracked yet. Mallory keeps watching.

VALID 0 / 0 TOTALView more in app

No public exploit code observed for this vulnerability.

EXPOSURE SURFACE

Affected products & vendors

Products and vendors Mallory has correlated with this vulnerability. Open in Mallory to drill down to specific CPE configurations and version ranges.

VendorProductType
Guardrails AIGuardrails Aiapplication

Vendor-confirmed product mapping. Mallory continuously reconciles this list against your asset inventory.

ACTIVITY FEED

Recent activity

8 sources tracked across advisories and community write-ups. News coverage will land here when it surfaces.

No news coverage yet. Advisories and community discussion only.

What this page doesn’t show

The version that knows your environment.

This page is what’s public. Mallory adds the parts that aren’t: which of your assets are affected, which adversaries are exploiting it right now, which detections to deploy, and what to do tonight.
Exposure mapping

Query your assets running an affected version, and investigate the blast radius.

Threat actor evidence

Every observed campaign linking this CVE to a named adversary.

Associated malware

Malware families riding this exploit, with evidence and IOCs.

Detection signatures

YARA, Sigma, Snort, and vendor rules, auto-deployed to your SIEM.

Vendor-by-vendor mapping

Cross-references every affected SKU, including bundled OEM variants.

Social activity8

Community discussion across Reddit, Mastodon, and other social sources.