The Agent Skills Directory

[SAFE]: The skill follows standard operational patterns for a content generation assistant. It reads a local voice profile and a structured JSON file of hook templates, both of which are expected for its stated purpose.
[SAFE]: Data storage is handled responsibly; the skill records its history to a local log file (~/.claude/linkedin/log.md) and maintains user voice profiles in a dedicated directory. This is standard practice for personalized AI agents.
[SAFE]: The skill includes explicit safeguards against common AI risks, such as a strict 'Never publish' rule and instructions to never fabricate data. It also correctly handles outbound links by advising they be placed in comments rather than the post body to avoid platform suppression.

li-post