competitor-teardown

Pass

Audited by Gen Agent Trust Hub on Jul 7, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references installation scripts for the belt CLI and additional specialized skills from the inference-sh platform's official GitHub repository.
  • [COMMAND_EXECUTION]: The skill utilizes various tools through the belt CLI to perform web searches, extract content from URLs, and capture screenshots for visual teardowns.
  • [COMMAND_EXECUTION]: Provides a Python script snippet using matplotlib to generate a positioning map visualization. The script is executed via the platform's python-executor tool.
  • [SAFE]: The skill ingests untrusted data from external sources such as competitor websites and search engines to perform analysis. This is a standard capability for research skills and does not involve executing the retrieved content.
  • [SAFE]: All referenced infrastructure and tools belong to the established inference-sh and belt-sh ecosystem, which is used here for its intended purpose of extending agent capabilities.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 7, 2026, 11:41 PM
Security Audit — agent-trust-hub — competitor-teardown