Skill Defender
Scans installed OpenClaw skills for malicious patterns including prompt injection, credential theft, data exfiltration, obfuscated payloads, and backdoors. Use when installing new skills, after skill updates, or for periodic security scans. Runs deterministic pattern matching — fast, offline, no API cost.
Savant verdict: Pending analysis
Imported from the source hub; Savant's analysis is still running.
Live evaluation
Not evaluated yet. Workspaces can request a live evaluation.
Safety (NVIDIA SkillSpector)
Scan pending.
SKILL.md
This listing is enumerated but its package hasn't been fetched yet; it's queued by popularity.