The Agent Skills Directory

[COMMAND_EXECUTION]: Uses the studio CLI to manage local WordPress development environments, including site lifecycle and configuration.
[INDIRECT_PROMPT_INJECTION]: Ingests data from external sites during imports and preview management. The skill provides clear instructions for the agent to treat this content as untrusted input.
Ingestion points: Site import via URL (--from) and preview site management described in SKILL.md.
Boundary markers: SKILL.md includes specific security callouts for the agent to handle untrusted user-generated content safely.
Capability inventory: File system operations, site deletion, and PHP evaluation via wp eval across CLI commands.
Sanitization: Relies on the agent following the provided safety instructions when processing site content.
[DYNAMIC_EXECUTION]: Enables PHP execution via studio wp eval and automated site setup using Blueprint JSON files. The skill in SKILL.md explicitly warns against using untrusted blueprints which could execute malicious code.
[CREDENTIALS_UNSAFE]: Manages WordPress administrator credentials in SKILL.md with explicit warnings to the agent against logging or caching passwords, and ensures they are omitted from machine-readable outputs by default.
[EXTERNAL_DOWNLOADS]: Fetches site data from user-specified URLs and retrieves internal tool updates (Data Liberation, Static Site Importer) which are checksum-verified by the vendor's platform.

studio-cli