The Agent Skills Directory

[COMMAND_EXECUTION]: The skill inherently requires executing shell commands and AppleScript to configure dev environments and automate macOS applications. It provides detailed guidance on using osascript securely, specifically advising on passing arguments via argv to prevent shell injection vulnerabilities.
[EXTERNAL_DOWNLOADS]: The documentation includes instructions for bootstrapping environments, which involves downloading well-known tools from trusted community sources.
Evidence: References the installation of Homebrew via its official site and the vim-plug manager from the reputable junegunn GitHub repository.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function involves scanning and auditing existing configuration files (dotfiles). This ingestion of external data represents a potential attack surface where malicious instructions could be embedded in user configs.
Ingestion points: The SCAN phase reads existing shell, editor, and terminal configuration files to understand the current environment.
Boundary markers: The skill relies on structured auditing patterns to detect anomalies rather than using explicit prompt delimiters for the file content.
Capability inventory: The skill can perform file writes, execute shell commands, and control system events on macOS.
Sanitization: It mitigates risk by proactively scanning for secrets using Gitleaks and auditing files for known security anti-patterns (e.g., hardcoded credentials or unsafe Git helpers).

hearth