The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill instructions direct the agent to ingest data from untrusted external sources (website URLs and Figma files) to perform branding tasks. This creates a surface where malicious content on a target website could attempt to influence the agent's behavior during the brand extraction and profile generation process.
Ingestion points: External website URLs and Figma URLs are passed to the create_or_update_brand_kit tool for scraping and extraction.
Boundary markers: The instructions do not specify any delimiters or safety markers to isolate external content from the agent's primary instructions.
Capability inventory: The skill uses create_or_update_brand_kit (modifies brand state) and apply_brand_kit_to_site (applies styling/configuration changes to a site).
Sanitization: There are no instructions for sanitizing or validating the content scraped from the provided URLs before processing.

replo-branding