The Agent Skills Directory

[PROMPT_INJECTION]: The skill uses detailed instructions to define a specific persona and writing style. It includes a self-check system to remove 'AI-sounding' patterns, which is a stylistic feature and does not attempt to bypass safety filters or ignore system-level constraints.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill uses local file storage for research notes and knowledge management, which is appropriate for its stated purpose. It does not access sensitive system files or credentials. All external references are to well-known services like Reddit, YouTube, and GitHub.
[EXTERNAL_DOWNLOADS]: The skill is installed via a legitimate author-controlled repository. No piped remote code execution or unauthorized downloads were detected.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from web searches and user-provided transcripts. However, it lacks dangerous capabilities like shell command execution or network exfiltration. The ingestion is confined to content generation and style transformation, following the primary purpose of the skill.
[DYNAMIC_CONTEXT_INJECTION]: No instances of load-time shell command execution were found in the instructions.

mj-writer