openclaw-security-monitor

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.80). The skill's SKILL.md/README explicitly states the IOC auto-updater (scripts/update-ioc.sh) fetches threat intelligence from an upstream GitHub repo and those fetched IOCs are used by the scanner/remediation logic (and the dashboard/server.js invokes remediation scripts), so public, potentially untrusted GitHub content can be ingested at runtime and directly influence scans and automated remediation actions.

MEDIUM W013: Attempt to modify system services in skill instructions.

• Attempt to modify system services in skill instructions detected (high risk: 0.90). The skill includes remediation scripts that modify file permissions, write to /etc/hosts, alter systemd/LaunchAgents and other system configs and can auto-apply fixes (--yes), effectively instructing agents to change system state and perform actions that may require sudo and can compromise the host if misused.