The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill's documentation (SKILL.md) provides standard installation instructions for prerequisites, including the use of sudo apt install python3 for Debian-based systems. This is a routine administrative command for environment setup and is appropriate for the skill's primary purpose.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided search queries to retrieve relevant design guidelines from its local datasets.
Ingestion points: The search.py script accepts a user query as a command-line argument.
Boundary markers: The output is structured with markdown headers, and the design_system.py script includes explicit disclaimers stating that the keyword matches are advisory and implementation remains unverified.
Capability inventory: The scripts are limited to local file reads of the provided CSV data and standard text processing. No network operations, file-write capabilities, or arbitrary command execution are present in the search logic.
Sanitization: The core.py script tokenizes the query by removing non-alphanumeric characters, and design_system.py cleans the output by stripping control characters from project names.
[SAFE]: The skill operates entirely locally, using the Python standard library for all processing. No external dependencies are required or installed at runtime, and no network calls are made by the core logic.

ui-ux-pro-max