/

CVE-2026-76216

Severity:

HIGH RISK

(7.5)

Authorization Bypass (CWE-639) in Vikunja LinkSharing IDs

CVE-2026-76216

Deferred

Vulnerability type

Category:

Authorization Bypass

CWE:

CWE-639

Impact:

Unauthorized access and modification of team and user resources

Severity:

HIGH RISK

(7.5)

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

Vikunja through version 2.4.0 contains a principal-type confusion vulnerability in permission checks involving LinkSharing principals. An attacker with a link-sharing JWT may cause a link-share identifier to be interpreted as a user identifier, potentially enabling unauthorized team changes, bot-user enumeration or deletion, and team-roster disclosure when identifiers collide.

Affected context

The vulnerability occurs in Vikunja permission checks that process LinkSharing principals and compare their identifiers with user identifiers without consistently validating the principal type. Risk is present when an attacker possesses a valid link-sharing JWT and identifier values overlap in the relevant autoincrement spaces.

Exploit preconditions

  • Attack surface: Vikunja team and user-management permission checks
  • Required access: A valid link-sharing JWT with permissions reaching an affected operation
  • Trigger point: Authorization decisions involving link-share or user principals
  • Why exploitation is possible: Missing principal-type guards allow a LinkSharing identifier to be treated as a user identifier when numeric IDs collide

Why this happens

The underlying bug is a failure to preserve and validate principal type during authorization decisions. Permission checks use an identifier without confirming whether it belongs to a LinkSharing principal or a user principal, allowing a value associated with one identity class to select an object in another class. This maps to CWE-639: Authorization Bypass Through User-Controlled Key, because an attacker-controlled or attacker-influenced key affects access control without adequate object and type validation.

Potential impact

  • An attacker may remove victims from teams
  • Victim bot users could be enumerated or deleted
  • Team rosters may be disclosed to unauthorized link-share users
  • Team membership and automation resources could be disrupted
  • Conflicting principal identifiers could cause broader authorization errors

General mitigation guidance

Mitigation generally focuses on making authorization decisions type-aware and binding every identifier to the intended principal class.

Common defensive patterns include:

  • Enforcing explicit principal-type checks before resource lookup and authorization
  • Using strongly typed identity representations rather than bare numeric identifiers
  • Binding authorization decisions to the authenticated principal and intended resource type
  • Applying deny-by-default behavior when identity type or ownership is ambiguous
  • Limiting link-sharing privileges to the minimum required operations
  • Avoiding reliance on identifier uniqueness across unrelated principal namespaces
  • Adding regression tests for identifier collisions and cross-type access attempts

How teams detect issues like this

Organizations can look for this class of issue through:

  • Static analysis rules that flag authorization lookups using untyped or user-controlled identifiers
  • Code review checks for missing principal-type or resource-type validation
  • Tests that exercise collisions between link-share, user, bot, and other identity namespaces
  • Policy enforcement requiring authorization checks to occur after identity and resource type resolution
  • Audit-log monitoring for unusual team removals, bot-user deletions, or roster access through link-sharing sessions
  • Automated security review tools that identify cross-object authorization flows

Sources

SourceLink
github.comgithub.com/go-vikunja/vikunja/security/advisories/GHSA-32r8-5843-4qw2
vulncheck.comvulncheck.com/advisories/vikunja-through-principal-type-confusion-via-linksharing

Frequently asked

questions

It may be exploitable through Vikunja's exposed application functionality when an attacker has a valid link-sharing JWT and the affected authorization paths are reachable.

Not necessarily. The relevant prerequisite is possession of a link-sharing token with access to an affected operation; the token's permitted scope and deployment configuration influence exposure.

Preserve principal types throughout authentication and authorization, validate the expected type before lookup, and deny requests when the identity-to-resource relationship is ambiguous.

Use typed identity models, test cross-namespace identifier collisions, apply least privilege to link-sharing tokens, and monitor sensitive actions performed through shared links.

Find out what's already
exploitable in your codebase.