OpenClaw allowlist bypass via PowerShell encoded-command aliases

Executive brief

OpenClaw, an AI assistant platform, contains a security flaw where its command filtering system can be bypassed. An attacker can use shortened versions of PowerShell commands to run unauthorized code that should have been blocked by the system's safety list. This could allow an unauthorized user to execute arbitrary commands on the host system, potentially leading to full system compromise or data theft.

Technical details

A vulnerability in OpenClaw's command validation logic allows for an allowlist bypass when processing PowerShell commands. The allowlist parser fails to recognize abbreviated or aliased forms of the '-EncodedCommand' flag (such as '-e', '-enc', or '-encodedc'). An attacker with low privileges can submit a command request using these unrecognized aliases to execute base64-encoded PowerShell payloads that bypass security policy checks. This issue is categorized as CWE-184 (Incomplete List of Disallowed Inputs). A fix is available in version 2026.5.12.

Affected products

• openclaw openclaw <= 2026.5.7

Timeline

• 2026-05-28: disclosed
• 2026-07-02: advisory: GitHub Advisory published
• 2026-05-12: patched

References

• https://api.github.com/users/YLChen-007
• https://github.com/YLChen-007
• https://api.github.com/users/YLChen-007/gists%7B/gist_id%7D
• https://api.github.com/users/YLChen-007/repos
• https://avatars.githubusercontent.com/u/30854794?v=4
• https://api.github.com/users/YLChen-007/events%7B/privacy%7D