claude-command-builder

Pass

Audited by Gen Agent Trust Hub on Jun 27, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is primarily instructional and serves to assist users in configuring their local environment with custom command shortcuts. It does not perform any autonomous network operations or access sensitive data outside of its documented scope.
  • [COMMAND_EXECUTION]: The skill demonstrates the use of the Claude Code bash execution syntax (!command). It correctly educates the user on the importance of the allowed-tools frontmatter to implement a least-privilege security model for these commands.
  • [INDIRECT_PROMPT_INJECTION]: As a template generator, the skill ingests user input (descriptions, arguments) to create code files. While the resulting commands could theoretically be manipulated by a malicious user, the skill itself provides validation steps in Step 2 to ensure command names and arguments follow safe naming conventions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 27, 2026, 11:07 AM
Security Audit — agent-trust-hub — claude-command-builder