Security auditing skill that detects malicious patterns like prompt injection, data exfiltration, obfuscation, and privilege escalation in OpenClaw SKILL.md...
Initial release of OSTRTA: a security auditing skill for OpenClaw skill files. - Provides adversarial analysis of SKILL.md files to detect prompt injection, data exfiltration, obfuscation, and other security threats. - Delivers clear verdicts (CRITICAL/HIGH/MEDIUM/LOW/SAFE) with evidence and remediation recommendations. - Supports optional cleaning/remediation: can generate a "safe" version with threats removed. - Detects 9 detailed threat categories including prompt injection, credential/data exfiltration, code obfuscation, unverifiable dependencies, and privilege escalation. - Uses an "assume-malicious" posture and presents actionable, evidence-based security findings.