Back to skill
Skillv1.0.0
VirusTotal security
LLM Switcher · External malware reputation and Code Insight signals for this exact artifact hash.
Scanner verdict
SuspiciousApr 30, 2026, 4:58 AM
- Hash
- 308a28768d70883c4cbef635764a48df08b7e22f95475f527474cbe2ceb316e8
- Source
- palm
- Verdict
- suspicious
- Code Insight
- Type: OpenClaw Skill Name: llm-switcher Version: 1.0.0 The skill's stated purpose is legitimate (changing LLM configuration and restarting the gateway). However, the instructions in SKILL.md for updating the default model (`openclaw config set agents.defaults.model.primary '"<selected-model>"'`) introduce a potential shell injection vulnerability. The `<selected-model>` value is derived from user input. While the skill instructs the agent to validate the model name against configured options, it does not explicitly instruct on sanitizing or escaping this input before embedding it into a shell command. An attacker could craft a malicious model name containing shell metacharacters to execute arbitrary commands on the host system.
- External report
- View on VirusTotal