CVE-2026-53845

Description

### Summary Skill-command dispatch could skip before-tool-call hooks. In affected versions, a skill command routed through the affected dispatch path could run without the same `runBeforeToolCallHook` coverage as other tool entry points. This advisory is scoped to the named feature and configuration. It does not change OpenClaw's trusted-operator model: authenticated Gateway operators, installed plugins, and intentional local execution surfaces remain trusted unless a separate policy, approval, allowlist, sandbox, or auth boundary is crossed. ### Impact When the affected feature is enabled and reachable, this could miss hook-based auditing or policy parity for that command path. Practical impact depends on the operator's configuration and whether lower-trust input can reach that path. ### Patched Versions The first stable patched version is `2026.5.6`. ### Mitigations avoid relying on hook-only enforcement for skill commands until patched. As general hardening, keep channel and tool allowlists narrow, avoid sharing one Gateway between mutually untrusted users, and disable the affected feature when it is not needed.

How to fix CVE-2026-53845

To remediate CVE-2026-53845, upgrade the affected package to a fixed version below.

• —upgrade to 2026.5.6 or later

Is CVE-2026-53845 being exploited?

No exploitation signal available. Neither CISA KEV nor a current EPSS score has been published for CVE-2026-53845.

Affected packages (1)

• from 0, < 2026.5.6

CVSS scores

References (4)

• ADVISORYnvd.nist.gov/vuln/detail/CVE-2026-53845
• PATCHgithub.com/openclaw/openclaw
• WEBgithub.com/openclaw/openclaw/security/advisories/GHSA-68xw-r643-9p5w
• WEBwww.vulncheck.com/advisories/openclaw-skill-command-dispatch-hook-bypass-via-before-tool-call-hook-skipping