The Agent Skills Directory

[PROMPT_INJECTION]: The SKILL.md file contains instructions labeled as 'CRITICAL RULE' using authoritative language ('MUST') to mandate a specific research workflow. This pattern is designed to override the agent's default decision-making process.
[INDIRECT_PROMPT_INJECTION]: The skill mandates that the agent scrape external UI galleries (such as Awwwards and ThreeUI) to extract design tokens. This creates a significant ingestion surface for untrusted data that could influence subsequent agent actions.
Ingestion points: Web scraping of external galleries via a multimodal subagent (specified in SKILL.md).
Boundary markers: None present to distinguish between instructions and ingested data.
Capability inventory: The skill possesses file writing capabilities and command execution via subprocess.run (found in scripts/shadcn_add.py).
Sanitization: No sanitization or validation of the scraped visual or textual content is implemented.
[COMMAND_EXECUTION]: The script scripts/shadcn_add.py uses subprocess.run to execute the npx CLI for component installation. While it avoids using a shell, it executes processes based on arguments derived from agent or user input.
[EXTERNAL_DOWNLOADS]: The skill instructions and scripts facilitate the download and execution of external components from the NPM registry using the npx shadcn@latest command.

ui-styling