Executive brief
OpenClaw, a tool used for managing shell commands and gateway operations, contains a security flaw in how it processes inline commands. An authenticated user could bypass security restrictions to run unauthorized commands that should have been blocked by an allowlist. This could allow an attacker with low-level access to execute arbitrary code or perform unapproved actions on the system.
Technical details
A vulnerability in OpenClaw (versions prior to 2026.5.12) allows for an allowlist bypass during the parsing of shell inline-command forms. The root cause is a specific parser case that fails to invoke the expected authorization logic, allowing commands to route around security checks. An authenticated attacker with network access can exploit this to execute shell content without the required approval prompts or allowlist validation. This issue is tracked as CVE-2026-53866 and was addressed in version 2026.5.12.
Affected products
- OpenClaw openclaw < 2026.5.12
Timeline
- 2026-05-28: advisory: Original GHSA-f397-5vjw-v2c2 published
- 2026-06-16: disclosed: NVD publication of CVE-2026-53866
- 2026-06-18: patched: Version 2026.5.12 released/confirmed as fix