The Agent Skills Directory

[SAFE]: The skill defines a structured workflow for analyzing video and visual references. It does not perform any system modifications, access sensitive files, or execute commands.
[NO_CODE]: The skill is implemented as a markdown file containing only instructions and frontmatter configuration. No scripts (Python, JavaScript, etc.) or binary files are included.
[EXTERNAL_DOWNLOADS]: The skill contains a reference to the developer's official website (postplus.io). This is an informational link provided in the documentation and does not involve automated code downloads or runtime execution.
[PROMPT_INJECTION]: No patterns of prompt injection, instruction overrides, or safety filter bypasses were identified in the instructions.
[COMMAND_EXECUTION]: The disable-model-invocation: true flag in the frontmatter explicitly prevents the agent from calling external tools while this skill is active, providing an additional layer of security.

reference-decode