The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill documents mechanisms for importing and rendering Bricks templates, which act as ingestion points for untrusted external data.
Ingestion points: The bricks/template/after_import action and the \Bricks\Frontend::render_data function mentioned in SKILL.md are designed to process and display data from potentially remote or third-party template sources.
Boundary markers: The code snippets do not provide instructions or implementations for boundary markers to distinguish between system instructions and data stored within template elements.
Capability inventory: The skill utilizes wp_insert_post and update_post_meta to store content, and \Bricks\Frontend::render_data to render that content, which could include malicious scripts if the source is compromised.
Sanitization: The documentation lacks explicit guidance or examples for sanitizing element settings or validating the structure of imported template JSON data.

bricks-templates