The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill's installation instructions in SKILL.md include commands that require elevated privileges to manage system packages (sudo apt update && sudo apt install python3). This directs the agent to attempt operations that require administrative rights.
[COMMAND_EXECUTION]: The skill's primary functionality relies on the execution of local Python scripts (scripts/search.py and scripts/design_system.py) to search databases and generate design recommendations. The workflow instructs the agent to run these scripts in the local environment to fulfill user requests.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied keywords through its Python scripts to retrieve AI prompt templates and design guidelines. This creates a potential surface where specially crafted user input could influence the instructions retrieved from the local database.
Ingestion points: User-controlled strings passed as arguments to the search.py script.
Boundary markers: The skill does not define explicit delimiters or protective instructions when presenting the search results to the agent.
Capability inventory: The included scripts perform local file-read operations (CSV) and output results to the terminal; they do not possess network access or file-writing capabilities.
Sanitization: The core.py script includes a tokenization process that uses a regular expression ([^\w\s]) to strip punctuation and non-alphanumeric characters from the input query, providing a basic layer of protection against traditional injection characters.

ui-ux-pro-max