The Agent Skills Directory

[COMMAND_EXECUTION]: The get_simulator_console and get_simulator_network tools allow the agent to process simulator logs using grep-style command strings. While intended for log analysis, this exposes a structured command interface for interacting with runtime data.
[COMMAND_EXECUTION]: The project-config module allows the agent to modify the project.config.json file. Since this file can contain pre-processing and post-processing shell scripts (e.g., beforeCompile), this represents a potential surface for indirect execution of local build scripts if the configuration is influenced by untrusted data.
[REMOTE_CODE_EXECUTION]: The automation_evaluate tool enables the execution of JavaScript code within the WeChat Mini-Program runtime context. This is a core functionality for UI automation and runtime debugging within the simulator environment.

wechatide-skill