The Agent Skills Directory

[SAFE]: The skill provides educational documentation and best practices for developing slash commands. It explicitly instructs developers to follow the principle of least privilege by using restrictive allowed-tools settings and provides guidance on securing commands against misuse.
[COMMAND_EXECUTION]: The documentation describes the use of the !command`` syntax for dynamic context gathering in slash commands. All examples provided in the skill and its references are pedagogical, using standard development tools like git, npm, and kubectl to illustrate intended functionality.
[INDIRECT_PROMPT_INJECTION]: The skill addresses the potential for prompt injection via user-supplied arguments and file references. It provides a dedicated "Validation Patterns" section that teaches developers how to implement sanitization, existence checks, and validation logic to safely handle untrusted data.
[SAFE]: Deterministic detections for potentially risky commands in references/testing-strategies.md were evaluated and determined to be benign. The dd and rm commands are used in a standard testing context to manage temporary files in /tmp, and other shell commands are part of negative test cases designed to verify platform-enforced security boundaries.

command-development