Summary
Argo Rollouts dashboard deployments through version 1.10.0 may expose rollout-changing operations on every network interface without authentication, authorization, or CSRF protection. An attacker with network access to the dashboard could alter, pause, restart, undo, or otherwise manipulate rollouts across namespaces available through the operator's kubeconfig.
Affected context
The issue occurs when the Argo Rollouts dashboard is exposed through a network-reachable listener that binds to all interfaces and does not require authenticated or authorized requests. Risk is greater when the dashboard process uses a kubeconfig with permissions spanning multiple namespaces, because dashboard operations may inherit that operational scope.
Exploit preconditions
- Attack surface: Network-accessible Argo Rollouts dashboard
- Required access: Network connectivity to the dashboard; valid user authentication may not be required
- Trigger point: Requests invoking mutating rollout operations
- Why exploitation is possible: The dashboard accepts state-changing actions without authentication, authorization, or CSRF protections and may execute them using the operator's kubeconfig permissions
Why this happens
The underlying issue is missing authentication and authorization on administrative API operations, mapping to CWE-306: Missing Authentication for Critical Function. The dashboard exposes mutating functionality without first establishing the caller's identity, verifying permission for the requested namespace and rollout, or requiring a CSRF-resistant request context. Binding the service to all interfaces can further expand exposure beyond trusted local clients.
Potential impact
- Attackers could alter rollout images or deployment progression
- Services could be aborted, restarted, undone, or repeatedly retried
- Applications may experience outages, degraded availability, or unexpected releases
- Rollout state and deployment integrity could be compromised across accessible namespaces
- Broad kubeconfig permissions could increase the scope of unauthorized changes
General mitigation guidance
Mitigation generally focuses on placing administrative dashboards behind authenticated and authorized access controls rather than exposing mutating APIs directly to untrusted networks.
Common defensive patterns include:
- Require strong authentication and explicit, namespace-scoped authorization for every mutating operation
- Bind administrative services to trusted interfaces or place them behind a hardened gateway
- Enforce CSRF-resistant request handling where browser-based access is supported
- Use dedicated, least-privileged service identities and narrowly scoped kubeconfig permissions
- Separate read-only observability access from rollout administration
- Apply network policies, allowlists, and secure ingress controls to limit dashboard reachability
- Prefer secure-by-default configurations that do not expose administrative listeners broadly
How teams detect issues like this
Organizations can identify this class of issue through:
- Static and configuration analysis for listeners bound to all interfaces without an authentication layer
- Code review checks for mutating handlers that lack identity, authorization, and CSRF validation
- API inventory reviews that distinguish read-only endpoints from state-changing operations
- Kubernetes and network policy audits for publicly or broadly reachable administrative dashboards
- Runtime monitoring for unexpected rollout mutations, image changes, aborts, restarts, and undo actions
- Policy enforcement requiring authenticated ingress and least-privilege service accounts
Automated security review tools may help flag unauthenticated administrative routes and overly broad network exposure.

