Executive brief
OpenClaw, a tool used for managing and executing remote commands, contains a flaw in how it displays pending tasks to administrators. An attacker with basic access can submit a very long command that appears harmless in the approval screen because the malicious part of the command is cut off from view. If an administrator approves the visible, benign-looking portion, the entire hidden malicious command will execute, potentially leading to a full system compromise or data theft.
Technical details
A User Interface (UI) Misrepresentation vulnerability (CWE-451) exists in OpenClaw's 'exec' approval workflow. The vulnerability is caused by improper handling of oversized command strings in the approval display component, which truncates long strings without indicating that more text exists. An authenticated attacker with permission to request command execution can craft a payload with a long, benign prefix followed by a malicious suffix. When an administrator reviews the request, they only see the benign prefix; however, upon approval, the backend executes the full, untruncated command string. This allows for unauthorized command execution with the privileges of the execution environment. The issue is addressed in version 2026.5.18.
Affected products
- OpenClaw OpenClaw < 2026.5.18
Timeline
- 2026-05-28: advisory: GitHub Security Advisory published
- 2026-06-12: disclosed: NVD publication date
- 2026-05-18: patched: First stable patched version released