The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The SKILL.md file contains setup instructions that suggest the use of high-privilege commands to install Python on Linux-based systems.
Evidence: In the 'Prerequisites' section, the skill recommends running sudo apt update && sudo apt install python3 if Python is not found.
[COMMAND_EXECUTION]: The skill's primary workflow involves the agent executing local Python scripts to query the design system databases.
Evidence: SKILL.md instructs the agent to run commands such as python3 .claude/skills/ui-ux-pro-max/scripts/search.py "<keyword>" --design-system to generate recommendations.
[INDIRECT_PROMPT_INJECTION]: The skill ingests user-supplied keywords and interpolates them into search queries that retrieve content from local CSV files, which is then processed by the agent.
Ingestion points: User requests are parsed into keywords for the search.py script in SKILL.md.
Boundary markers: The skill does not explicitly instruct the agent to use boundary delimiters when interpreting the output of the local search scripts.
Capability inventory: The skill relies on local script execution and file reading (csv, json) via Python.
Sanitization: The scripts/core.py script implements basic sanitization by tokenizing input and removing non-word characters (re.sub(r'[^\w\s]', ' ', str(text).lower())).

ui-ux-pro-max