The Agent Skills Directory

[COMMAND_EXECUTION]: The scripts/format.py script uses subprocess.Popen to open the local system browser for viewing the theme gallery. Additionally, scripts/publish.py uses subprocess.run to call the formatting script internally. These are standard operations for local development tools and do not pose a security risk as they do not process unsanitized shell commands from external sources.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user-supplied Markdown articles. The instructions in SKILL.md explicitly direct the agent to only add structural markers (like titles and callouts) without altering the user's original wording or following instructions embedded within the article text, which serves as a protective boundary against adversarial input.
[EXTERNAL_DOWNLOADS]: The skill depends on standard Python packages (markdown, requests, Pillow) and communicates with official WeChat API endpoints (api.weixin.qq.com) and AI service providers (e.g., OpenRouter) to perform its functions. These network operations are necessary for article publishing and content generation and do not involve fetching untrusted code for execution.

hkr-render