The Agent Skills Directory

[SAFE]: The skill consists entirely of documentation, tutorials, and examples for command development. It does not contain any malicious code, obfuscation, or unauthorized data access patterns.\n- [DYNAMIC_CONTEXT_INJECTION]: The documentation describes and provides examples for using the !command syntax to execute shell commands. These examples are properly contained within markdown code blocks and do not trigger execution during skill loading. The skill provides clear guidance on using restrictive allowed-tools filters (e.g., Bash(git:*)) to mitigate security risks in user-created commands.\n- [COMMAND_EXECUTION]: The skill includes numerous examples of shell commands for common development tasks such as code linting, testing, and deployment. These are provided as template patterns for users to adapt and do not involve any suspicious or unknown external sources or active execution during the skill's operation.\n- [INDIRECT_PROMPT_INJECTION]: The skill explicitly addresses risks associated with processing untrusted input in slash commands. It provides several patterns for argument validation, schema verification, and resource existence checks to help developers build secure integrations and prevent injection attacks.

command-development