The Agent Skills Directory

[PROMPT_INJECTION]: The skill's primary workflow involves retrieving and processing content from external websites, creating a surface for indirect prompt injection.
Ingestion points: SKILL.md (Workflow 1) and references/design-resources.md instruct the agent to browse and analyze data from third-party sites like Dribbble, Mobbin, Behance, and Awwwards.
Boundary markers: The instructions lack delimiters (e.g., XML tags or clear separators) and specific guidance for the agent to ignore or isolate embedded instructions within the processed web content.
Capability inventory: The skill utilizes powerful tools including chrome-devtools (which can read network/console data), ai-multimodal (Gemini API for analysis), and mentions shell-based tools like gallery-dl, FFmpeg, and ImageMagick via the media-processing and bash integrations.
Sanitization: There is no evidence of filtering or sanitization of the data retrieved from external URLs before it is processed by the agent's LLM.

aesthetic