The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides instructions to download and execute a shell script directly from a remote GitHub repository by piping it to a shell (curl -fsSL https://raw.githubusercontent.com/jackwener/wx-cli/main/install.sh | bash). This pattern allows for the execution of arbitrary, unverified code from an external source.
[COMMAND_EXECUTION]: The skill requires the user to run sudo wx init to scan system memory and extract database encryption keys. Executing third-party tools with root privileges is a significant security risk.
[COMMAND_EXECUTION]: The skill instructions involve force-signing the WeChat application (`codesign --force --deep --sign
/Applications/WeChat.app) and resetting the macOS Transparency, Consent, and Control (TCC) database (tccutil reset`). These actions modify application integrity and system-level security configurations.
[EXTERNAL_DOWNLOADS]: The skill downloads and installs external software from the NPM registry (@jackwener/wx-cli) and a remote GitHub repository, which introduces supply chain risks.
[DATA_EXFILTRATION]: The primary function of the skill is to access and extract sensitive private data from local WeChat databases, including chat history, contact lists, and encryption keys.
[PROMPT_INJECTION]: The skill processes untrusted data from WeChat messages and search results (via commands like wx history and wx search). This creates a surface for indirect prompt injection, where malicious content in a message could attempt to influence the agent's logic. The skill lacks boundary markers or sanitization procedures for this external content.

wx-cli