The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The SKILL.md file contains prerequisite setup instructions that direct the agent to execute privileged commands using sudo on Linux environments (e.g., sudo apt update && sudo apt install python3).
[COMMAND_EXECUTION]: The skill operates by executing local Python scripts (scripts/search.py) to process user requirements and retrieve design recommendations from its internal database.
[EXTERNAL_DOWNLOADS]: The data files contain numerous references and URLs for external resources, including Google Fonts and official documentation for frameworks like Next.js, Flutter, and shadcn/ui. These references target well-known services and trusted organizations.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user input by interpolating search queries into a script that returns text from a design database back to the agent, potentially allowing malicious content within the database to influence agent behavior.
Ingestion points: User-provided search queries passed as arguments to scripts/search.py.
Boundary markers: The skill does not implement explicit delimiters or instructions to ignore embedded commands in the retrieved data.
Capability inventory: The skill utilizes local shell execution for Python scripts.
Sanitization: The query is tokenized in scripts/core.py using regular expressions to remove punctuation and filter short words, which provides a basic layer of input sanitization.

ui-ux-pro-max