The Agent Skills Directory

[COMMAND_EXECUTION]: The skill extensively uses Node.js child_process modules to facilitate browser automation and system interaction.
It spawns Google Chrome instances with debugging ports enabled for CDP automation in cdp.ts and wechat-browser.ts.
It executes platform-specific commands for clipboard management, including osascript and swift on macOS, powershell on Windows, and xclip/wl-copy on Linux in copy-to-clipboard.ts.
It uses spawnSync to orchestrate internal markdown rendering processes and agent-browser for alternative automation flows.
[EXTERNAL_DOWNLOADS]: The skill facilitates the download of external resources necessary for content processing.
md-to-wechat.ts fetches remote images from user-provided URLs to prepare them for WeChat upload.
utils/languages.ts dynamically loads syntax highlighting grammars from a well-known CDN (OSS Aliyun).
[REMOTE_CODE_EXECUTION]: The documentation and diagnostic scripts reference a common piped installation pattern.
SKILL.md and check-permissions.ts suggest the command curl -fsSL https://bun.sh/install | bash to help users install the Bun runtime.
Because this targets a well-known service and requires manual execution by the user during setup, it is classified as a safe instructional reference rather than an automated vulnerability.
[DATA_EXFILTRATION]: The skill transmits content and credentials to external servers as part of its primary purpose.
wechat-api.ts sends article data and media to api.weixin.qq.com.
These transmissions are directed exclusively to official WeChat endpoints and are required for the skill to function as intended.

baoyu-post-to-wechat