The Agent Skills Directory

[PROMPT_INJECTION]: The skill creates an indirect prompt injection surface by generating a configuration file that serves as a 'source of truth' for other skills based on external data. * Ingestion points: Data is collected from the mcp__motion__get_workspace_brand, mcp__motion__get_workspace_competitors, and get_glossary_values tools (triggered in SKILL.md), as well as direct user input. * Boundary markers: None are present; user and API data are interpolated directly into the configuration template without delimiters or instructions to ignore embedded commands. * Capability inventory: The skill possesses Write and Edit capabilities (SKILL.md) to modify the local file system. * Sanitization: There is no evidence of sanitization or validation of the ingested strings before they are persisted to the configuration file.
[COMMAND_EXECUTION]: The skill uses file-system tools (Read, Write, Edit) to manage its own configuration file (motion-creative.config.md) and brief templates. This is standard operational behavior for a setup skill.

customize