Description

A flaw was found in Red Hat Quay's JWT (JSON Web Token) validation for federated robot accounts and single sign-on (SSO) authentication. Multiple issues related to audience verification and the enforcement of `azp` and `sub` claims were identified. These flaws could allow an attacker with a validly-signed token from the same identity provider to bypass configured security restrictions. This bypass could lead to unauthorized access by circumventing intended audience, subject, or authorized-client limitations.

Executive Summary

Technical Details

Mitigations

References

Additional Metadata

{
  "score": "5.4",
  "pubDate": "2026-08-14T23:16:33.960Z",
  "pubdate": "2026-08-14T23:16:33.960Z",
  "executiveSummary": "A validation flaw exists within Red Hat Quay's JSON Web Token (JWT) verification mechanism affecting federated robot accounts and single sign-on (SSO) authentication. The vulnerability stems from improper audience verification and insufficient enforcement of the authorized party (azp) and subject (sub) claims.\nAn external attacker possessing a validly-signed token issued by the same identity provider can exploit these deficiencies to circumvent strict security boundaries. By bypassing intended audience, subject, and authorized-client limitations, unauthorized entities can successfully authenticate and gain unauthorized access to protected resources within the Quay environment.\nThe risk implication is a compromise of access control policies enforced via federated mechanisms. Exploitation requires the attacker to already hold a validly-signed token from the trusted identity provider, which is then crafted or reused to trick the Quay validation logic into accepting out-of-scope claims.\nAffected systems include Red Hat Quay instances utilizing federated robot accounts or SSO authentication mechanisms relying on the flawed JWT parsing and claim verification routines.",
  "technicalDetails": "The root cause of this vulnerability lies in the cryptographic and logical validation implementation of JSON Web Tokens within Red Hat Quay's authentication pipeline. Specifically, the component responsible for processing tokens for federated robot accounts and single sign-on (SSO) authentication fails to adequately enforce mandatory claims and constraints.\nDuring the JWT validation lifecycle, cryptographic verification ensures the token was signed by a trusted identity provider (IdP). However, the subsequent claims validation stage fails to rigorously verify the audience (aud) claim, the authorized party (azp) claim, and the subject (sub) claim against the expected context of the requested operation or service endpoint.\nBecause of these missing or weak validations, an attacker who obtains a legitimately signed JWT from the shared identity provider—even one intended for a completely different application, client, or audience—can present this token to Red Hat Quay. The vulnerable component evaluates the signature as authentic and improperly trusts the accompanying claims without enforcing strict scope matching.\nThe step-by-step attack flow proceeds as follows: First, the attacker acquires a validly-signed token from the identity provider. Second, the attacker transmits this token to the Red Hat Quay authentication endpoint supporting federated robot accounts or SSO. Third, Quay's JWT validation routines successfully verify the cryptographic signature but omit or miscalculate the verification of azp, sub, and aud claims. Finally, Quay grants an authenticated session or authorizes access based on the accepted token, allowing the attacker to bypass intended access control restrictions and impersonate or access unauthorized resources.\nThe vulnerable component involves the authentication and token-processing logic handling federated robot accounts and SSO integrations within Red Hat Quay. Authentication requirements for the initial token depend on the IdP, but exploitation permits unauthorized access within Quay without requiring administrative privileges on the target system, provided the IdP token is acquired."
}