The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses the node:child_process module's spawnSync function to execute the WeChat DevTools CLI (cli or cli.bat). This is a necessary and standard procedure for enabling automation ports and managing the development tool environment.
[DATA_EXPOSURE]: The skill provides templates for taking screenshots of the mini-program simulator using miniProgram.screenshot(). This is an expected feature for visual regression and E2E testing.
[PROMPT_INJECTION]: The skill sets up a surface for indirect prompt injection by listening to logs and exceptions generated by the mini-program under test.
Ingestion points: Runtime console output and exceptions are ingested into the agent context via miniProgram.on('console') and miniProgram.on('exception') in SKILL.md.
Boundary markers: No specific boundary markers or instructions are provided to the agent to ignore potentially malicious content within these logs.
Capability inventory: The skill has access to command execution (spawnSync) and file system operations (fs.mkdir, fs.writeFile via the library's screenshot capability).
Sanitization: Log data is processed without sanitization before being evaluated by the agent.

miniprogram-automation