The Agent Skills Directory

[PRIVILEGE_ESCALATION]: The skill instructions advise users to use sudo for installing system-level dependencies required for the search scripts.
Evidence: sudo apt update && sudo apt install python3 in SKILL.md (line 365).
[INDIRECT_PROMPT_INJECTION]: The skill ingests extensive data from bundled CSV files and presents it to the agent, creating a vulnerability surface where instructions hidden in the data could influence agent behavior.
Ingestion points: The core search logic in scripts/core.py and scripts/design_system.py reads data from numerous files including data/styles.csv, data/prompts.csv, and various CSV files in the data/stacks/ directory.
Boundary markers: The results extracted from the CSV databases are displayed to the agent without specific delimiters or instructions to ignore embedded commands.
Capability inventory: The skill is configured with powerful tools including claude-code, codex-cli, and opencode, which allow the agent to execute shell commands and modify project files based on its input.
Sanitization: There is no evidence of content sanitization or validation for the data retrieved from the CSV files before it is loaded into the agent's context.

ui-ux-pro-max-v2