The Agent Skills Directory

[SAFE]: The skill is a pure reference library (Pattern 10) with its mode explicitly set to 'knowledge' and 'user-invocable' set to false. It serves as an educational repository for architecture principles without containing executable logic or scripts.
[EXTERNAL_DOWNLOADS]: The documentation includes links to official specifications and well-known technology resources (e.g., docs.oracle.com, maven.apache.org, docs.claude.com). These are standard references for the technical domain covered by the skill.
[COMMAND_EXECUTION]: While various files (e.g., 'skill-patterns.md' and 'reference-patterns.md') contain code blocks demonstrating bash or python commands, these are strictly illustrative examples intended for developers learning how to build their own skills. They are not designed to be executed by the skill itself.
[PROMPT_INJECTION]: The 'execution-directive.md' file provides guidance on using mandatory markers (e.g., MANDATORY, CRITICAL) to improve LLM adherence to instruction workflows. This is a documented prompt engineering technique for development and does not attempt to subvert the host agent's safety protocols.
[CREDENTIALS_UNSAFE]: No hardcoded credentials, API keys, or secrets were detected. The documentation correctly references placeholders like '' in examples and promotes the use of environment variables for secret management.

plugin-architecture