agent-folder-init

Pass

Audited by Gen Agent Trust Hub on Apr 10, 2026

Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes a local Python script (scaffold.py) to generate the folder structure. This script performs file system operations such as creating directories and writing template files to the project root.
  • [COMMAND_EXECUTION]: The skill generates numerous slash command templates (e.g., /analyze-codebase, /clean, /validate) that instruct AI agents to perform shell operations like tree, grep, find, and git for codebase analysis and maintenance.
  • [PROMPT_INJECTION]: The skill scaffolds specialized agent persona files (e.g., senior-backend-engineer.md) that contain 'MANDATORY READING' instructions. These instructions use authoritative framing to steer the agent's behavior and enforce project-specific coding standards.
  • [SAFE]: The scaffolding script includes a directory traversal check (is_relative_to(cwd)) to ensure it does not write files outside the intended project directory without explicit user consent via the --allow-outside flag.
  • [SAFE]: No network exfiltration, remote code downloads, or multi-layered obfuscation patterns were detected in the source code or templates.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 10, 2026, 09:45 AM