The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses validation scripts (validate.sh and validate.ps1) that utilize dynamic execution functions like eval and Invoke-Expression to run commands.
[COMMAND_EXECUTION]: The commands to be executed are dynamically extracted from local project files such as package.json (e.g., the lint and test scripts). This creates an attack surface where a repository with a malicious configuration could execute arbitrary shell commands when the agent triggers the validation loop.
[COMMAND_EXECUTION]: The skill is intended for autonomous, "fire-and-forget" operation, which means these validation commands are executed without a human-in-the-loop to verify the safety of the project-specific scripts being run.

apex-tier1