agent-tools

Pass

Audited by Gen Agent Trust Hub on Aug 31, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill downloads the 'belt' CLI binary and installation scripts from cli.inference.sh and dist.inference.sh. These are the official domains for the inference.sh platform, which is the primary provider for this skill's functionality.
  • [REMOTE_CODE_EXECUTION]: The installation instructions prominently feature curl -fsSL https://cli.inference.sh | sh. While this pattern executes remote code, it is a documented method for installing the platform's official tooling. The skill also provides a safer manual alternative involving binary verification via SHA-256 checksums, which is a security best practice.
  • [COMMAND_EXECUTION]: The skill makes extensive use of the belt CLI via Bash commands to perform tasks such as generating images, running LLMs, and uploading files. This is the intended and primary purpose of the skill. The CLI also includes commands for managing shell completions, which involves writing to standard shell configuration paths (e.g., /etc/bash_completion.d/).
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 31, 2026, 11:00 PM
Security Audit — agent-trust-hub — agent-tools