skills/cosmix/loom/loom-before-after/Gen Agent Trust Hub

loom-before-after

Pass

Audited by Gen Agent Trust Hub on Jul 25, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill is purely educational and procedural, providing structural guidance for creating verification stages in a development workflow. It focuses on proving that features are added or bugs are fixed by comparing system state before and after an implementation stage.
  • [COMMAND_EXECUTION]: The skill's allowed tools include Bash, which is used in examples to run standard testing and diagnostic commands such as cargo test, curl, jq, and grep. These commands are scoped to the local development environment (e.g., localhost:8080) and are appropriate for the skill's stated purpose of verifying code changes.
  • [CREDENTIALS_UNSAFE]: The examples demonstrate how to test authenticated endpoints using placeholders like Bearer <token> or Bearer fake. No real secrets, API keys, or credentials are hardcoded in the skill.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 25, 2026, 07:34 AM
Security Audit — agent-trust-hub — loom-before-after