CVE-2026-15318: MQTT Authorization Bypass in Sipeed PicoClaw

Technical summary

CVE-2026-15318 is an authorization bypass vulnerability in the MQTT Channel Handler component (pkg/channels/mqtt/mqtt.go) of Sipeed PicoClaw. The flaw stems from improper validation of the client_id argument, allowing authenticated users to manipulate this parameter and circumvent access controls. The vulnerability maps to CWE-285 (Improper Authorization) and CWE-863 (Incorrect Authorization), indicating that the application fails to properly enforce access restrictions based on user identity or role. The attack surface is network-accessible, requires valid authentication credentials, and produces no known user interaction requirement.

Business impact

This authorization bypass enables authenticated users to access MQTT channels and functionality beyond their intended permissions. In deployments where PicoClaw is used for IoT device management or industrial automation, an insider or compromised account could potentially read sensitive telemetry, modify device configurations, or disrupt operations. The moderate CVSS score reflects the requirement for prior authentication, but the ease of exploitation and public availability of working exploit code elevates practical risk in active threat environments.

Affected systems

Sipeed PicoClaw versions up to and including 0.2.9 are affected. This tool is commonly used in embedded systems development, particularly in Rust-based IoT and edge computing projects. Organizations relying on PicoClaw for remote device management, firmware deployment, or MQTT-based control systems should prioritize inventory and assessment.

Exploitability

The vulnerability requires an attacker to possess valid MQTT credentials, but the actual exploitation is trivial once authenticated. Public exploit code is available, removing the barrier to weaponization. The low attack complexity and absence of special conditions mean that any authenticated user—whether internal, compromised, or a supply-chain threat—can immediately attempt the attack. The lack of KEV inclusion suggests limited evidence of active in-the-wild exploitation at time of CVE publication, but the public code release materially increases risk trajectory.

Remediation

Update Sipeed PicoClaw to a patched version beyond 0.2.9. Verify the specific patch version against the official Sipeed repository and release notes. In parallel, implement network segmentation to restrict MQTT access to trusted hosts, enforce strong authentication on MQTT brokers, and monitor for suspicious client_id patterns in MQTT logs. If immediate patching is not feasible, restrict MQTT listener access to internal networks only.

Patch guidance

Check the Sipeed PicoClaw GitHub repository for releases after version 0.2.9. Apply the latest stable release to all affected deployments. Test patches in a non-production environment first, particularly if PicoClaw is integrated with critical IoT or automation workflows. Verify that MQTT channel authorization is correctly enforced post-patch by testing access control with different user roles and client IDs.

Detection guidance

Monitor MQTT broker logs for unusual client_id values, especially ones that appear forged or inconsistent with your naming conventions. Look for authenticated sessions accessing channels outside their normal scope. Network detection rules should flag repeated failed authorization attempts followed by successful access, which may indicate exploitation attempts. Application logs in pkg/channels/mqtt should be reviewed for any authorization warnings or access denials that correlate with suspicious activity.

Why prioritize this

Although the CVSS score is moderate, the combination of public exploit code, ease of exploitation post-authentication, and the network-accessible nature of MQTT brokers elevates this to a near-term remediation priority. Any organization using PicoClaw in production, especially for device management or critical automation, should treat this as high priority. The lack of KEV inclusion does not diminish the practical risk given public code availability.

Risk score, explained

The CVSS:3.1 score of 6.3 (MEDIUM) reflects a network-accessible vulnerability requiring prior authentication (PR:L). However, this score does not fully capture the ease of exploitation once credentials are obtained or the prevalence of overly-permissive MQTT broker configurations in IoT deployments. The score appropriately penalizes the authentication requirement but does not heavily weight the public exploit availability, which in practice makes this a faster-moving risk than the 'MEDIUM' label might suggest to busy security teams.

Frequently asked questions