The Agent Skills Directory

[SAFE]: No security issues detected. The skill follows a structured 'human-in-the-loop' design pattern, necessitating user confirmation at every step of the media search and retrieval process. This prevents the agent from autonomously calling tools or modifying code without oversight.- [PROMPT_INJECTION]: The skill does not contain any instructions that attempt to override safety guidelines, bypass system constraints, or extract sensitive instructions. The provided rules strictly define the intended workflow for media retrieval.- [DATA_EXFILTRATION]: While the skill involves handling URLs from Salesforce CMS and Data Cloud, these are processed as legitimate data assets for user-approved application features. The instructions specifically guide the agent on how to correctly use these URLs (including necessary query parameters) for their intended purpose. No patterns of unauthorized data transfer or access to sensitive local configuration files were found.- [COMMAND_EXECUTION]: The skill facilitates the use of specific internal search tools (search_media_cms_channels and search_electronic_media) for a defined and restricted purpose. The workflow enforces that no tools are called until the user has explicitly chosen a search method.

searching-media