Shell Environment Variable Injection in OpenClaw Exec Handling

OpenClaw versions 2026.2.22 through 2026.4.12 suffer from insufficient validation of shell-wrapper detection, enabling attackers to inject malicious environment variable assignments at the argument vector level. By bypassing exec preflight checks, adversaries can manipulate critical shell variables such as SHELLOPTS and PS4 to alter execution semantics and circumvent security controls. This vulnerability affects any organization deploying vulnerable OpenClaw versions in production environments where shell command execution is involved, particularly those relying on OpenClaw for secure process isolation or command sanitization. With a CVSS score of 8.8, this represents a high-severity risk that could lead to privilege escalation, command injection, or security policy bypass.

While this CVE currently has zero mapped MITRE ATT&CK techniques and no matching Casky skills in our catalog, practitioners using Casky's Claude-powered platform with extended reasoning would detect attack patterns through behavioral anomaly detection and environment variable manipulation signatures. Security teams should monitor for: (1) unexpected SHELLOPTS or PS4 variable modifications in process execution logs, (2) argv-level argument injection attempts that circumvent standard input validation, and (3) execution flows that deviate from normal OpenClaw wrapper behavior. As threat intelligence matures on this CVE, Casky's skill library will expand to include detection rules mapping to Execution and Defense Evasion techniques, enabling practitioners to correlate environment variable tampering with command execution anomalies in their security investigations.