The Agent Skills Directory

[SAFE]: The skill consists entirely of instructional markdown files and SQL/shell snippets designed to assist with database administration and optimization. It does not contain any hidden executable scripts, obfuscated code, or malicious logic.
[COMMAND_EXECUTION]: The skill documents various pscale CLI commands and administrative SQL queries (e.g., pg_terminate_backend, DROP INDEX, pscale branch delete). Crucially, the author has included proactive safety warnings throughout the documentation, explicitly instructing the AI agent to confirm with a human before executing any command that could be destructive to data or availability.
[CREDENTIALS_UNSAFE]: Static detectors flagged connection strings in the documentation. However, these are strictly templates or examples using obvious placeholders such as <user>, <password>, xxx, and pscale_pw_xxx. No actual live credentials, API keys, or private tokens are hardcoded in the skill files.
[EXTERNAL_DOWNLOADS]: The skill references official resources from PlanetScale's GitHub repository and suggests standard tools like the pscale CLI and the pg_activity monitoring utility. These are legitimate tools for the intended use case and originate from the vendor's official channels.

postgres