The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local Python scripts (publish.py, article_init.py, and getdraft.py) to automate interactions with the WeChat Official Account API and manage local article files.
[DATA_EXFILTRATION]: Accesses WeChat API credentials (APPID, APPSECRET) stored in aws.env and transmits them to api.weixin.qq.com (a well-known service) to facilitate authentication and content publishing. It also uploads article data and images to the same destination as part of its primary function.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes user-provided content from article.html and article.yaml for publishing.
Ingestion points: Reads content and metadata from article.html and article.yaml within article directories.
Boundary markers: The skill instructions define a "publish_completed gate" that requires checking for placeholder text within the content before marking a publication as complete.
Capability inventory: Includes shell command execution for article management, filesystem writes for metadata and links, and network POST requests to the WeChat API.
Sanitization: Content is checked for specific placeholders, but the skill otherwise relies on the agent or user for content safety before transmission.

aws-wechat-article-publish