Back to skill
Skillv0.5.1
VirusTotal security
Skill · External malware reputation and Code Insight signals for this exact artifact hash.
Scanner verdict
SuspiciousApr 30, 2026, 4:03 AM
- Hash
- b8b72736dc789bc94ba115aadb26239c90eb41050c1fc8d537bee830b6111714
- Source
- palm
- Verdict
- suspicious
- Code Insight
- Type: OpenClaw Skill Name: grazy Version: 0.5.1 The skill is classified as suspicious primarily due to clear prompt injection attempts in SKILL.md. The instructions repeatedly tell the AI agent to "IMMER zuerst `grazy help` aufrufen" (always call `grazy help` first) when uncertain. While the `grazy help` command itself is benign, this instruction is an attempt to influence the agent's decision-making and prioritize a specific action, which is a form of prompt injection. There is no evidence of direct malicious intent such as data exfiltration, persistence, or unauthorized remote control within the provided files, but the manipulation of agent behavior warrants a 'suspicious' classification.
- External report
- View on VirusTotal