The Agent Skills Directory

[PROMPT_INJECTION]: The skill exhibits a potential surface for indirect prompt injection by ingesting untrusted data from markdown files to generate executable HTML prototypes.
Ingestion points: Reads content from docs/{功能名称}/1_需求文档.md and specs/1_产品概述.md (as specified in SKILL.md).
Boundary markers: Absent; there are no explicit instructions or delimiters used to separate user data from instructions or to warn the agent to ignore embedded commands within the requirement documents.
Capability inventory: The agent has file-write capabilities, outputting generated HTML files to the docs/{功能名称}/prototypes/ directory.
Sanitization: Absent; the skill does not specify any validation, escaping, or filtering of the ingested markdown content before it is interpolated into the generated HTML/JavaScript code.
[EXTERNAL_DOWNLOADS]: The skill generates HTML files that include a script reference to the Tailwind CSS CDN (https://cdn.tailwindcss.com). This is a reference to a well-known service essential for the skill's primary prototyping functionality and is considered a safe reference to a standard technology provider.

ui-prototype