The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands for setup and functionality, including creating directories with mkdir -p, opening files with open -a Preview, and executing curl to interact with image generation APIs.
[EXTERNAL_DOWNLOADS]: The skill retrieves content from external websites for brand analysis and competitor research, and it communicates with external AI service providers like OpenAI and kie.ai.
[PROMPT_INJECTION]: The skill's architecture is vulnerable to indirect prompt injection. Ingestion points: Untrusted data enters the context via scraping of product URLs and customer reviews. Boundary markers: There are no explicit markers to separate scraped content from agent instructions. Capability inventory: The skill can perform shell command execution and file operations based on processed data. Sanitization: No sanitization is documented for the external content before it is used to generate ad concepts and prompts.

meta-ads-generator