Approval Policy Bypass in OpenClaw Skill Workshop

Detection Guidance

This vulnerability involves the Skill Workshop apply flow in OpenClaw versions before 2026.5.6, where an agent tool call can bypass the approval policy by setting apply: true despite approvalPolicy: pending.

Detection would involve monitoring or auditing calls to the Skill Workshop apply path to identify if any apply requests are being made with apply: true while approvalPolicy is still pending.

Suggested detection steps include:

• Review logs or audit trails for Skill Workshop apply requests that set apply: true without completing the approval step.
• Restrict and monitor agent tool calls that interact with the apply flow.
• Manually review Skill Workshop changes for unauthorized application.

No specific commands are provided in the available resources. However, general commands could include searching logs for relevant apply calls, for example using grep or similar tools on log files:

• grep -i 'apply: true' /path/to/openclaw/logs/*
• grep -i 'approvalPolicy: pending' /path/to/openclaw/logs/*

Further detection may require custom scripts or monitoring tools tailored to the OpenClaw environment and its logging mechanisms.

Useful 0 Not Useful 0

Executive Summary

This vulnerability exists in OpenClaw versions before 2026.5.6 and involves an approval policy bypass in the Skill Workshop apply flow.

Specifically, it allows agent tool calls to set the apply flag to true even when the approvalPolicy is still pending.

Attackers can exploit this flaw by accessing the affected apply path to make workshop changes before the required approval step has been completed.

This means configurations can be modified without proper authorization.

Useful 0 Not Useful 0

Impact Analysis

This vulnerability can impact you by allowing unauthorized changes to configurations within the OpenClaw Skill Workshop.

Because the approval step can be bypassed, attackers may modify settings or workflows without proper permission.

This could lead to potential misuse or disruption of the system's intended operations.

Useful 0 Not Useful 0

Compliance Impact

The vulnerability allows unauthorized modification of workshop configurations by bypassing the approval step, which could lead to improper authorization controls.

Such unauthorized changes may impact compliance with standards and regulations that require strict access controls and change management, such as GDPR and HIPAA.

However, the provided information does not explicitly discuss the direct impact on compliance with these or other common standards.

Useful 0 Not Useful 0

Mitigation Strategies

To mitigate the approval policy bypass vulnerability in OpenClaw before version 2026.5.6, you should:

• Manually review Skill Workshop changes to detect unauthorized modifications.
• Restrict the use of the agent tool that can set apply: true despite approvalPolicy: pending.
• Keep channel and tool allowlists narrow to limit exposure.
• Avoid sharing Gateways between untrusted users to reduce risk.
• Disable the affected Skill Workshop apply feature when it is not needed.

Additionally, upgrading to OpenClaw version 2026.5.6 or later, which contains the patch for this vulnerability, is recommended.

Useful 0 Not Useful 0