setup-borzoi-remote-tool

Pass

Audited by Gen Agent Trust Hub on Aug 24, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructs the user to install a Python package directly from a remote URL (https://connect.aiscientist.tools/downloads/tuplatform_connect-0.3.0-py3-none-any.whl). While the instruction includes a SHA-256 hash to ensure file integrity, the domain is not a globally recognized trusted service or a standard package registry like PyPI.
  • [REMOTE_CODE_EXECUTION]: The installation of a binary wheel from an external, non-whitelisted domain constitutes remote code execution, as the package content is executed upon installation. This is partially mitigated by the use of hash pinning for verification.
  • [COMMAND_EXECUTION]: The skill makes extensive use of shell-based command execution, including running local preflight scripts (scripts/remote_validation/setup_skill_preflight.py) and executing Python code blocks via heredoc redirection (`python
  • <<'PY'`). These scripts are used for tool discovery and scientific validation.
  • [DATA_EXFILTRATION]: The skill provides instructions for handling sensitive credentials such as TOOLUNIVERSE_SERVICE_KEY and TOOLUNIVERSE_API_TOKEN. It follows secure practices by advising the use of protected 0600 environment files, utilizing browser-based device authorization flows to avoid manual key pasting, and warning against passing secrets as shell arguments.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 24, 2026, 08:20 AM
Security Audit — agent-trust-hub — setup-borzoi-remote-tool