The Agent Skills Directory

[DATA_EXFILTRATION]: The skill includes an optional input parameter wechat_profile_dir that requests the path to a Chrome profile directory. Accessing browser profile directories is a high-risk activity because these folders contain sensitive data, including session cookies, authentication tokens, browser history, and potentially stored passwords.
[COMMAND_EXECUTION]: The workflow involves executing shell commands to run Node.js scripts from associated helper skills belonging to the same author (e.g., node <zhy-article-illustrator>/scripts/illustrate-article.ts, node <zhy-markdown2wechat>/scripts/convert.js, and node <zhy-wechat-publish>/scripts/wechat_draft.js). These commands are used to automate the full article lifecycle from illustration to publishing.
[INDIRECT_PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as it retrieves and processes content from external websites and search engine results without explicit sanitization or the use of boundary markers.
Ingestion points: External URLs provided via the urls input and content retrieved from the WebSearch and webfetch tools.
Boundary markers: The instructions do not define specific delimiters or instructions to the agent to disregard potential embedded commands within the fetched material.
Capability inventory: The skill possesses the capability to write files to the system and execute Node.js scripts via the shell.
Sanitization: No content filtering or sanitization procedures are described for the data ingested from the web.

zhy-wechat-writing