Executive brief
OpenClaw is a Node.js-based tool that enables operators to dispatch commands and run plugins. A bug in its skill-command dispatcher can allow authenticated users to bypass hook-based security controls that normally enforce auditing and policy checks. An attacker with authenticated access could run commands without triggering required validation hooks, potentially circumventing audit logs or compliance policies.
Technical details
This is a hook-bypass vulnerability (CWE-693) in OpenClaw's skill-command dispatch path. When skill commands are routed through a specific dispatch path, they skip the runBeforeToolCallHook coverage that other tool entry points enforce. The vulnerability requires an authenticated operator with Gateway access; it does not affect the core trusted-operator model. An attacker can send skill commands that bypass hook-based auditing and policy enforcement, though practical impact depends on the operator's configuration and whether lower-trust input can reach that path. The issue is patched in version 2026.5.6.
Affected products
- OpenClaw OpenClaw <=2026.5.5
Timeline
- 2026-05-28: disclosed
- 2026-05-28: patched: Version 2026.5.6 released
- 2026-06-16: advisory: GHSA-r7vv-6763-m739 published (later withdrawn as duplicate)
- 2026-06-18: other: GHSA-r7vv-6763-m739 withdrawn as duplicate of GHSA-68xw-r643-9p5w