GHSA-48WF-G7CP-GR3M: OpenClaw Allowlist Bypass via 'env -S'
Vulnerability ID: GHSA-48WF-G7CP-GR3M
CVSS Score: 8.8
Published: 2026-03-03
A critical security bypass exists in OpenClaw's execution guard mechanism, allowing attackers to circumvent binary allowlists using the env utility's split-string feature. By leveraging env -S, an attacker can execute arbitrary commands even when the system is configured to restrict execution to specific safe binaries. This vulnerability stems from a semantic mismatch between the policy engine's validation logic and the runtime behavior of command-line wrappers.
TL;DR
OpenClaw's allowlist can be bypassed using env -S to execute arbitrary commands. The fix distinguishes between transparent wrappers and semantic execution modifiers, blocking dangerous flags.
⚠️ Exploit Status: POC
Technical Details
- CWE ID: CWE-78
- Attack Vector: Local (via Agent)
- CVSS Score: 8.8
- Impact: Arbitrary Code Execution
- Exploit Status: PoC Available
- Remediation: Patch Available
Affected Systems
- OpenClaw (Moltbot/ClawdBot) < v2026.2.24
-
OpenClaw: < v2026.2.24 (Fixed in:
v2026.2.24)
Code Analysis
Commit: a1c4bf0
fix(security): harden exec wrapper allowlist execution parity
Commit: 3f923e8
test: add env -S allowlist bypass regressions
Mitigation Strategies
- Upgrade OpenClaw to version v2026.2.24 or later immediately.
- Audit existing allowlists to remove generic wrappers like
env,sh, orbashunless strictly necessary. - Implement strict argument validation for any allowlisted binary that accepts sub-commands.
Remediation Steps:
- Pull the latest version of the OpenClaw repository or Docker image.
- Verify the installation of the patch by checking for the
fix(security)commit hasha1c4bf07c6baad3ef87a0e710fe9aef127b1f606. - Review the
config.yamlor relevant policy file to ensuresafeBinsdoes not include unnecessary broad-scope binaries. - Restart the OpenClaw service to apply the new execution guard logic.
References
Read the full report for GHSA-48WF-G7CP-GR3M on our website for more details including interactive diagrams and full exploit analysis.
Top comments (0)