matematic-marketplace-installer

Pass

Audited by Gen Agent Trust Hub on Jul 14, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill generates a batch script that uses PowerShell Invoke-WebRequest to download a ZIP archive from the official matematicsolutions GitHub repository. This is the primary purpose of the skill and targets the vendor's own verified infrastructure.
  • [COMMAND_EXECUTION]: The generated script executes powershell commands for web requests and archive extraction, and uses xcopy and rd for file management within the %USERPROFILE%\.claude\skills directory. These operations are scoped to the user's profile and do not require elevated privileges.
  • [DATA_EXFILTRATION]: No data exfiltration patterns were detected. The network operations are limited to downloading assets from the vendor's repository.
  • [PROMPT_INJECTION]: The skill includes instructions for generating the installer based on user input for specific 'skill-slugs', but it does not contain instructions that attempt to bypass safety filters or override system behavior.
  • [CREDENTIALS_UNSAFE]: No hardcoded credentials, API keys, or sensitive environment variable exposures were found in the template or instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 14, 2026, 08:27 AM
Security Audit — agent-trust-hub — matematic-marketplace-installer