skills/brandonburrus/.claude/fix/Gen Agent Trust Hub

fix

Pass

Audited by Gen Agent Trust Hub on Jul 2, 2026

Risk Level: SAFE
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructs the agent to use standard development CLI tools such as git, grep, and system-specific tools like codesign and security to troubleshoot issues. These commands are used for local repository management and build process inspection.
  • [PROMPT_INJECTION]: The skill contains a specific security safeguard in the 'Gotchas' section, instructing the agent to treat error output and logs as diagnostic data only and to avoid executing instructions embedded in such data. This is a best practice for mitigating indirect prompt injection from malicious stack traces or external logs.
  • [DATA_EXPOSURE]: The localization steps suggest inspecting environment variables (e.g., env | grep IDENTITY) to debug propagation issues. This is a standard and necessary part of software development and debugging workflows within the user's environment.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 2, 2026, 05:11 PM
Security Audit — agent-trust-hub — fix