Back to skill
Skillv0.1.3
VirusTotal security
Lybic Sandbox · External malware reputation and Code Insight signals for this exact artifact hash.
Scanner verdict
SuspiciousApr 28, 2026, 8:18 AM
- Hash
- 4a63b834a088bfdafb2628069a1a055bde438109e2a857b2b2002d953a0e40be
- Source
- palm
- Verdict
- suspicious
- Code Insight
- Type: OpenClaw Skill Name: lybic-sandbox Version: 0.1.3 The skill provides extensive capabilities for arbitrary code execution, GUI automation (mouse, keyboard, touch), file system access (download, read, write), and network operations (port mapping) within a Lybic cloud sandbox. While these are high-risk capabilities, they are explicitly aligned with the stated purpose of controlling a 'disposable cloud computer' for automation workflows. There is no clear evidence of intentional harmful behavior targeting the host system, such as credential theft, data exfiltration from the host, persistence mechanisms, or malicious prompt injection against the agent to subvert its core directives. The instructions in SKILL.md and the example code (e.g., examples/01_execute_code.py, examples/03_file_processing.py, examples/06_http_port_mapping.py) demonstrate these powerful features, which, while necessary for the skill's function, inherently carry a high risk if misused by the agent or a malicious prompt.
- External report
- View on VirusTotal