/

CVE-2026-56249

Severity:

HIGH RISK

(7.6)

Authorization Bypass (CWE-285) in Capgo Channel Logic

CVE-2026-56249

Deferred

Vulnerability type

Category:

Authorization Bypass

CWE:

CWE-285

Impact:

Privilege Escalation and Configuration Tampering

Severity:

HIGH RISK

(7.6)

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

Capgo contains an authorization bypass in its channel creation endpoint that lets authenticated users overwrite existing channels by reusing their names.

When a user has app.create_channel permission, a logic mismatch between existence checks and upsert operations can reassign ownership and alter critical production channel configurations.

Affected context

This vulnerability arises in deployments where Capgo exposes an application API for creating and managing channels, and where channel names are used as a key for both existence checks and upsert operations.

It becomes relevant when users with app.create_channel permission can interact with production or shared channels, and when the system treats create operations as idempotent or upsert-like without re-validating ownership or access rights.

The issue is most pronounced in multi-tenant or team-based environments where channels are shared across accounts and are used to control production configuration or release flows.

Exploit preconditions

  • Attack surface: Application/API endpoint responsible for channel creation and management.
  • Required access: Authenticated account with app.create_channel permission, but without legitimate ownership of the target channel.
  • Trigger point: Submission of a channel creation request using the name of an already existing channel, leveraging the create endpoint's upsert behavior.
  • Why exploitation is possible: The system validates that a user is allowed to create a new channel but does not correctly enforce ownership and authorization when an existing channel with the same name is found, allowing the upsert to overwrite or reassign the channel.

Why this happens

The underlying bug is an authorization logic error where business rules for "create" operations are incorrectly applied to "update/upsert" scenarios.

Instead of separating the authorization paths for new channel creation and existing channel modification, the endpoint relies on a simplified existence check and then proceeds with an upsert-like operation without re-verifying that the caller is authorized to alter the existing resource.

This maps to CWE-285: Improper Authorization, where control over a sensitive resource (channels and their configuration) is granted based on incomplete or inconsistent permission checks and identity/ownership validation.

Potential impact

  • Attackers may reassign ownership of existing channels to their own account.
  • Attackers could modify critical production channel configurations.
  • Attackers can disrupt deployment and release flows that depend on channel settings.
  • Attackers may indirectly escalate privileges by gaining control over channels used for access or rollout gating.
  • Attackers could cause application downtime or misrouting of traffic due to misconfigured channels.

General mitigation guidance

Effective mitigation focuses on enforcing strict authorization boundaries between resource creation and modification.

Common secure design patterns include:

  • Separating "create" and "update" operations and applying distinct authorization checks for each path.
  • Ensuring that upsert-like operations explicitly validate ownership and access rights when an existing resource is detected.
  • Using immutable or strongly scoped identifiers for channels (e.g., IDs) and treating names as attributes, not as primary keys for authorization decisions.
  • Applying least-privilege principles so that app.create_channel permission does not imply the ability to modify or take over existing channels.
  • Incorporating server-side business rules that prevent reassignment of ownership without explicit privileged operations and audit.

How teams detect issues like this

Detection typically involves examining authorization flows and resource handling in the channel management code.

Helpful strategies include:

  • Static analysis or manual review focusing on endpoints that perform create, update, or upsert behavior, verifying that authorization checks differ by operation type.
  • Inspecting code for patterns where existence checks (e.g., "find or create") are not paired with ownership or role validation when a resource already exists.
  • Implementing policy-based access control frameworks (RBAC/ABAC) and verifying that resource-level checks are enforced consistently at the service layer, not only at the UI.
  • Monitoring logs for unusual channel ownership changes or frequent re-creations of existing channel names, which may indicate exploitation.

Sources

SourceLink
github.comgithub.com/Cap-go/capgo/security/advisories/GHSA-vj24-j594-3wv3
vulncheck.comvulncheck.com/advisories/capgo-unauthorized-channel-overwrite-and-ownership-takeover-via-post-channel-name-collision

Frequently asked

questions

Any authenticated user who holds the app.create_channel permission and can access the channel creation endpoint may be able to exploit the logic mismatch, even if they do not legitimately own the target channel.

Not necessarily. The vulnerability is triggered by a specific permission (app.create_channel), which may be granted to non-admin roles. The risk depends on how broadly this permission is assigned and how channels are used in production.

The most robust fix is to clearly separate creation and modification logic and enforce explicit resource-level authorization on updates. When a channel with the requested name already exists, the system should treat the request as a modification and verify ownership, roles, or higher-privilege permissions before allowing any change.

To prevent comparable issues, adopt a consistent access control model that:

  • Applies authorization checks at the resource level for all state-changing operations.
  • Avoids reliance on names or user-supplied identifiers for authorization decisions.
  • Uses comprehensive security tests for business logic paths, including "find or create" and upsert workflows.
  • Integrates automated review of access control code and periodic audits of permission assignments.
Find out what's already
exploitable in your codebase.