The Agent Skills Directory

[COMMAND_EXECUTION]: The skill includes Python scripts such as package_macos_app.py that utilize subprocess calls to execute standard development tools like lipo or llvm-lipo. These tools are used to merge binaries into universal macOS apps. The script implements these calls securely using argument lists and checks for the existence of the tools in standard system paths.
[EXTERNAL_DOWNLOADS]: The project configuration and documentation rely on official and well-known software delivery paths. This includes standard Go module dependency management and the use of official GitHub Actions for Continuous Integration (CI) and release automation.
[DATA_EXFILTRATION]: Analysis of the Go starter project and utility scripts shows no evidence of unauthorized data collection or exfiltration. Data persistence patterns in the code follow best practices, such as using system-defined configuration directories for game saves.
[OBFUSCATION]: All code, scripts, and instructions within the skill are provided in plain text. There is no evidence of Base64 encoding for commands, zero-width characters, or other techniques designed to hide malicious behavior.
[PROMPT_INJECTION]: The SKILL.md instructions provide clear and safe guidance for the AI agent. It does not contain any patterns intended to bypass safety filters, override system prompts, or induce unrestricted behavior.

ebitengine-game