Summary
@backstage/plugin-auth-backend: Unauthenticated OAuth account takeover via redirect_uri allowlist bypass
Workarounds
Disable the experimental features by removing auth.experimentalDynamicClientRegistration and auth.experimentalClientIdMetadataDocuments from your app-config, which is the default configuration. Alternatively, restrict the configured allowlist patterns to fully specified URLs with an explicit protocol and no wildcard in the hostname, which are not affected by this vulnerability.
Impact
The allowlist matching used by the experimental dynamic client registration and client ID metadata document (CIMD) features in @backstage/plugin-auth-backend matched glob patterns against the full URL string. A * wildcard could therefore match across URL component boundaries: a pattern such as https://*.example.com/callback, intended to allow subdomains of a trusted host, would also match an attacker-controlled URL such as https://attacker.example/x.example.com/callback. This applies to auth.experimentalDynamicClientRegistration.allowedRedirectUriPatterns as well as the allowedClientIdPatterns and allowedRedirectUriPatterns options of auth.experimentalClientIdMetadataDocuments.
An attacker could use this to register an OAuth client whose redirect URI points to a host they control while still passing the allowlist, causing authorization codes to be delivered to the attacker when a victim completes an authorization flow. In addition, allowlist patterns without an explicit protocol could match URLs with any protocol, and redirect URIs containing embedded credentials (user:pass@host) were accepted after the credentials were stripped for matching.
The practical impact is limited. Both features are experimental and disabled by default, and the default allowlist patterns only reference fixed or loopback hosts and are not affected. Deployments are only impacted if they enable one of these features and configure custom allowlist patterns that contain a wildcard in the hostname, or patterns without an explicit protocol.
Untrusted input controls a URL used for redirection, which can forward users to attacker-controlled sites. Typical impact: phishing and credential harvesting via a trusted domain.
GHSA-38HQ-7X33-PHP4 has a CVSS score of 4.7 (Medium). The vector is network-reachable, no privileges required, and user interaction required. A CVSS score reflects the worst-case severity of the vulnerability, not your specific exposure. Whether this affects your application depends on whether the vulnerable code is present and reachable in your environment. A fixed version is available (0.29.2); upgrading removes the vulnerable code path.
Affected versions
Security releases
Kodem intelligence
Severity tells you how bad this could be in the worst case. It does not tell you whether you are exposed. Exploitability and impact are functions of runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A vulnerable package can sit in your dependency tree and never run.
Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter. Kodem's runtime-powered SCA identifies whether this CVE is reachable in your applications.
Already deployed Kodem?
See it in your environmentNew to Kodem? Get a demo →Remediation advice
Patched in @backstage/plugin-auth-backend version 0.29.2. Patterns are now matched against each URL component separately so that wildcards no longer match across the host and path boundary, patterns without an explicit protocol are rejected as invalid configuration, and redirect URIs with embedded credentials are always rejected.
Note that as part of this fix, a wildcard port no longer implicitly matches every path: a pattern such as http://localhost:* now only matches the root path. Use http://localhost:*/* to allow any port and any path.
Frequently Asked Questions
- What is GHSA-38HQ-7X33-PHP4? GHSA-38HQ-7X33-PHP4 is a medium-severity open redirect vulnerability in @backstage/plugin-auth-backend (npm), affecting versions <= 0.29.1. It is fixed in 0.29.2. Untrusted input controls a URL used for redirection, which can forward users to attacker-controlled sites.
- How severe is GHSA-38HQ-7X33-PHP4? GHSA-38HQ-7X33-PHP4 has a CVSS score of 4.7 (Medium). This score reflects the worst-case severity of the vulnerability, not your specific exposure. Whether it represents real risk in your environment depends on whether the vulnerable code is present and reachable.
- Which versions of @backstage/plugin-auth-backend are affected by GHSA-38HQ-7X33-PHP4? @backstage/plugin-auth-backend (npm) versions <= 0.29.1 is affected.
- Is there a fix for GHSA-38HQ-7X33-PHP4? Yes. GHSA-38HQ-7X33-PHP4 is fixed in 0.29.2. Upgrade to this version or later.
- Is GHSA-38HQ-7X33-PHP4 exploitable, and should I be worried? Whether GHSA-38HQ-7X33-PHP4 is exploitable in your environment depends on whether the vulnerable code is present and reachable. A CVSS score is a worst-case rating; it does not account for your specific deployment, configuration, or usage patterns. Kodem, an Intelligent Application Security platform, uses runtime intelligence to show which vulnerabilities actually execute in production, so you can focus on the ones that represent real risk. Get a demo
- What actually determines whether GHSA-38HQ-7X33-PHP4 is exploitable, and how bad it is? Exploitability and impact are not fixed properties of a CVE. They depend on runtime truth: whether the vulnerable code is present, reachable, and actually executes in your application. A high CVSS score on a dependency that never runs is not the same as real risk. Kodem, an Intelligent Application Security platform, uses runtime intelligence to reveal which vulnerabilities actually execute in production, so teams prioritize the ones that genuinely matter.
- How do I fix GHSA-38HQ-7X33-PHP4? Upgrade
@backstage/plugin-auth-backendto 0.29.2 or later.