case-study-writing

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS] The skill references and provides instructions for installing the belt CLI tool from inference.sh and refers to installation scripts hosted on GitHub (raw.githubusercontent.com/inference-sh/skills/...). These are documented as vendor-provided tools for the skill's functionality.
  • [COMMAND_EXECUTION] The skill utilizes the belt CLI to execute various sub-commands (belt login, belt app run) to perform web searches and execute code. This is part of the primary functional design.
  • [DYNAMIC_EXECUTION] The skill includes an example of executing Python code via the infsh/python-executor app to generate charts using matplotlib. The code is provided as a static template within the instructions and does not ingest untrusted runtime data for execution.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 05:09 PM
Security Audit — agent-trust-hub — case-study-writing