/

CVE-2026-45549

Severity:

HIGH RISK

(8.5)

Missing Authorization Checks (CWE-862) in Roxy-WI Agent

CVE-2026-45549

Deferred

Vulnerability type

Category:

Missing Authorization

CWE:

CWE-862, CWE-863

Impact:

Privilege Escalation and Remote Service Control

Severity:

HIGH RISK

(8.5)

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

Roxy-WI versions 8.2.6.4 and earlier contain a missing authorization control that allows any authenticated user, including low-privilege guests, to trigger management actions on monitoring agents.

By calling a protected but under-authorized endpoint, an attacker can start, stop, or restart the roxy-wi-smon-agent systemd unit on arbitrary servers reachable by Roxy-WI over SSH, causing service disruption or abuse of root-level execution.

Affected context

The vulnerability occurs in deployments of Roxy-WI (8.2.6.4 and prior) where the agent_action endpoint is exposed to authenticated users via the web interface.

The affected route, /agent/action/<action>, is protected only by JWT-based authentication and does not enforce role-based access control or ownership checks on the server_ip parameter.

Environments where Roxy-WI connects to managed servers using SSH with passwordless sudo for the roxy-wi-smon-agent systemd unit are particularly exposed, as actions are executed with root privileges on those targets.

Exploit preconditions

  • Attack surface: Roxy-WI web interface endpoint POST /agent/action/<action>.
  • Required access: Any valid, authenticated Roxy-WI account, including lowest-privilege (role 4 / guest).
  • Trigger point: Submission of a crafted POST request specifying:
  • A permitted action value (e.g., start/stop/restart).
  • A server_ip corresponding to a server managed by Roxy-WI.
  • Why exploitation is possible: The route uses @jwt_required() for authentication but lacks:
  • Role-based authorization checks to restrict who can perform agent actions.
  • Group or ownership checks to verify that the caller is allowed to manage the specified server_ip.
  • Enforcement of least privilege on the SSH/sudo configuration, causing actions to run as root on the target.

Why this happens

The root cause is incomplete authorization logic on a sensitive administrative endpoint, mapping to CWE-862: Missing Authorization and related CWE-863: Incorrect Authorization.

While the endpoint enforces authentication via JWT, it does not implement fine-grained permission checks to ensure only privileged roles can invoke agent actions, nor does it limit operations to servers the user is authorized to manage.

This design assumes that all authenticated users can safely access high-impact functionality, violating separation of duties and least-privilege principles and effectively turning a low-privilege session into a conduit for root-level actions on remote hosts.

Potential impact

  • Remote starting, stopping, or restarting of the roxy-wi-smon-agent service on arbitrary managed servers.
  • Disruption of monitoring or management workflows by stopping agents across the fleet.
  • Indirect impact on availability of load-balanced or monitored services if operational visibility is degraded.
  • Potential use of agent restarts as a stepping stone for further misconfiguration or exploitation of other weaknesses in the environment.

General mitigation guidance

Effective mitigation focuses on enforcing strict, role-aware authorization and narrowing the trust boundary for remote operations:

  • Introduce explicit role-based access control (RBAC) on the agent_action route, granting access only to administrative or operator roles.
  • Enforce ownership or group membership checks on server_ip so users can act only on servers assigned to them or their groups.
  • Apply the principle of least privilege to SSH and sudo configurations used by Roxy-WI, avoiding blanket passwordless root privileges where possible.
  • Centralize access control decisions in reusable authorization helpers or middleware to reduce the risk of unprotected administrative endpoints.
  • Consider adding server-side policy rules that define which actions are allowed per role and per server, and have the application enforce these policies consistently.

How teams detect issues like this

Organizations can detect and monitor for this class of issue using a combination of code-level and operational techniques:

  • Static analysis and code review focused on identifying routes that:
  • Are protected only by authentication decorators (e.g., @jwt_required()), and
  • Invoke privileged operations (systemd control, SSH commands) without role or ownership checks.
  • Security review patterns that flag any endpoint performing remote execution or service control as requiring explicit RBAC and resource-scoped authorization.
  • Runtime logging and monitoring of agent_action requests, including user identity, role, target server_ip, and action, with alerts on anomalous patterns (e.g., guests controlling many servers).
  • Use of automated application security testing tools or CI pipelines to enforce policies such as "all administrative routes must invoke authorization helpers" and to detect deviations from those patterns.

Sources

SourceLink
github.comgithub.com/roxy-wi/roxy-wi/security/advisories/GHSA-c92j-h72m-ff4j

Frequently asked

questions

It can be, if the Roxy-WI web interface is exposed to the internet and an attacker can obtain or create any authenticated account. The exploit does not require direct network access to the target servers, only to Roxy-WI itself.

Any authenticated Roxy-WI account is sufficient, including the lowest-privilege "guest" role. No additional administrative privileges are required because the missing authorization checks allow guests to invoke high-impact actions.

Implement robust role-based authorization and resource ownership checks on all agent and server-management routes. Ensure that only explicitly privileged roles can trigger system-level actions on managed servers and that those actions are restricted to resources the user is allowed to control.

Adopt a security-by-design approach where every endpoint that touches remote execution, system services, or infrastructure management must:

  • Use centralized authorization helpers or middleware.
  • Define clear, testable policies for which roles can perform which actions on which resources.
  • Be covered by automated tests and code review checklists that verify proper authorization logic before deployment.
Find out what's already
exploitable in your codebase.