The Agent Skills Directory

[SAFE]: The skill's primary purpose is project organization and maintainability auditing. It defines clear decision rules that prioritize framework constraints and project-native conventions.
[SAFE]: The bundled script scripts/structure-snapshot.mjs is designed for passive reconnaissance of folder structures. It explicitly avoids reading file contents to maintain privacy.
[SAFE]: The script contains a robust blacklist of sensitive file names and directories (e.g., .env, .secrets, credentials, .git) to prevent accidental exposure of credentials or private metadata during structural scans.
[SAFE]: The instructions mandate 'Approval Gates,' requiring the agent to ask for user permission before making changes that affect public APIs, deployment configurations, secrets, or large data sets.
[SAFE]: The skill recommends using .env files for secret management, which is a standard security best practice.
[SAFE]: No external downloads, remote code execution patterns, or data exfiltration vectors were detected. All analysis and script execution are performed locally on the user's project.

project-folder-structure