Skip to content
Submit incident
AII-1001-09417
ConfirmedA court, a regulator or the organization itself has stated the core facts, or the record cites a court or regulator source.

OpenClaw node host system.run rawCommand/command mismatch allows allowlist bypass

OpenClaw's node host system.run handler contained a mismatch between the rawCommand parameter used for allowlist checks and the command[] array used for actual execution. This enabled attackers with system.run access to bypass security controls by providing an approved command name while executing a different command through the argv parameter.

15 February 2026 Β· openclaw Β· Security and cyberattacks
15 February 2026
AII-1001-09417View source β†—
Quick facts
Organization
OpenClaw (subject)
Category
Security and cyberattacks
Source
github.com
Date
15 February 2026
Drafted by
Raidu Incident Detection

What happened

A mismatch between rawCommand and command[] in the OpenClaw node host system.run handler could cause allowlist and approval evaluation to be performed on one command while executing a different argv.

An attacker with the ability to invoke system.run could bypass allowlist enforcement and approval prompts by supplying an allowlisted rawCommand while providing a different command[] argv for execution.

The vulnerability affected deployments using the node host execution path with allowlist-based exec policy and approval prompting enabled.

Reported impact

Attackers could execute commands that should have been blocked or required user approval by exploiting the inconsistency between the allowlist-checked command and the actual executed command.

Index analysis

The flaw was a validation gap where the system checked permissions against one command string but executed a different command specified in an array parameter.

Governance controls

OpenClaw published a patch in version 2026.2.14 that enforces rawCommand and command[] consistency through gateway fail-fast validation and node host validation.

Classification

Organization
OpenClaw (subject)
Category
Security and cyberattacks
Harm status
Potential harm
Intent
Unintentional
Timing
Post-deployment
AI systems
OpenClaw (openclaw)

Sources and evidence

This record was researched and written by the Index, based on the following original reporting.

github.com
https://github.com/openclaw/openclaw/security/advisories/GHSA-h3f9-mjwj-w476
Aug 2026Claude AI agent discovers gym booking system vulnerability, cancels another user's reservationWell documentedSecurity and cyberattacksThe IndependentDiaspora Digital Media (DDM News) - Nigeria Breaking News, Africa and World News and Updates -The Indian Express+3Jun 2026Five malicious skills removed from OpenClaw ClawHub marketplaceConfirmedSecurity and cyberattacksUnit42Dark ReadingJun 2026OpenClaw AI agent vulnerabilities enable malicious skill packages to steal user dataUnder investigationSecurity and cyberattacksenglish.news.cnChina NewsApr 2026AI Agent Exploited Gym Booking API Authorization Flaw and Removed Member From WaitlistReportedSecurity and cyberattacksabc.net.auaffinda.comApr 2026Open-source AI agents expose security vulnerabilities in ChinaReportedSecurity and cyberattacksδΈœζ–Ήθ΄’ε―Œη½‘China News