/

CVE-2026-72564

Severity:

CRITICAL RISK

(9.6)

Improper Authorization (CWE-639) in Pangolin Resources

CVE-2026-72564

Deferred

Vulnerability type

Category:

Improper Authorization

CWE:

CWE-639

Impact:

Cross-resource unauthorized access and authentication bypass

Severity:

CRITICAL RISK

(9.6)

CVSS version:

3.1

Catch this CVE while code is being written

Detect vulnerable code directly in pull requests, so CVEs are fixed before merge.

Summary

An improper authorization vulnerability in Pangolin allows an authenticated remote attacker to reuse an access token issued for one resource to authenticate to a different resource in any organization. The issue becomes risky when resource and organization authorization checks do not sufficiently bind tokens to their intended scope.

Affected context

This issue may affect Pangolin deployments through version 1.20.0 when access tokens are accepted across resource or organization boundaries without adequate scope validation. It is relevant to authenticated remote access flows that rely on bearer tokens to authorize requests for protected resources.

Exploit preconditions

  • Attack surface: Remote resource authentication and authorization endpoints
  • Required access: A valid authenticated access token issued for at least one resource
  • Trigger point: Authentication or authorization of a request targeting a different resource
  • Why exploitation is possible: The authorization process may fail to verify that the token is bound to the requested resource and organization

Why this happens

The underlying bug is an authorization design failure in which a credential issued for one protected object is treated as valid for another object without enforcing object-level ownership or scope restrictions. This maps to CWE-639: Authorization Bypass Through User-Controlled Key, because resource selection or identification can be accepted without confirming that the authenticated principal is authorized for the selected resource.

Potential impact

  • An authenticated attacker may access resources outside their authorized scope
  • The attacker could view or interact with protected resource functionality
  • Organization and tenant boundaries may be weakened
  • Sensitive data or operational actions may be exposed
  • The issue can enable broader account or environment compromise depending on resource permissions

General mitigation guidance

Mitigation generally focuses on binding every authorization decision to the authenticated identity, token scope, requested resource, and organization context.

Common defensive patterns include:

  • Enforcing server-side object-level authorization for every resource request
  • Validating that token claims and scopes match the requested resource and organization
  • Avoiding authorization decisions based solely on user-controlled resource identifiers
  • Applying least-privilege scopes and minimizing token permissions by default
  • Using short-lived, audience-restricted tokens with explicit resource binding
  • Separating tenant and organization authorization checks from general authentication

How teams detect issues like this

Organizations can look for this class of issue through:

  • Static analysis and code review checks for missing object-level authorization
  • Review of token validation logic, especially resource, audience, scope, and tenant claims
  • Tests that attempt to reuse credentials across resources and organizations
  • API policy enforcement requiring an authorization decision for every protected object
  • Audit-log analysis for successful access to resources outside a principal's normal scope

Automated security testing and code review tools can help identify authorization paths that validate identity but omit resource ownership checks.

Sources

SourceLink
github.comgithub.com/fosrl/pangolin
github.comgithub.com/fosrl/pangolin/blob/main/server/routers/resource/authWithAccessToken.ts

Frequently asked

questions

Yes. The described issue can be exercised remotely through resource authentication or authorization flows, provided the attacker has a valid token.

Yes. An attacker generally needs an access token issued for a different resource, but does not necessarily need legitimate access to the target resource.

Bind each authorization decision to the authenticated principal, requested resource, organization, and token scope, then deny access when any relationship is not explicitly valid.

Use object-level authorization checks on every request, apply least-privilege token scopes, isolate organization contexts, and include cross-resource and cross-tenant authorization tests in security reviews.

Find out what's already
exploitable in your codebase.