The Agent Skills Directory

[SAFE]: The skill implements a drafting workflow for social media content using internal reference templates (e.g., hook-formulas.md, voice-profile.md).
[SAFE]: All network operations, specifically the publishing of content via lib.publish, are gated by a mandatory manual approval step (Step 8), which mitigates the risk of unauthorized data exfiltration or automated posting of malicious content.
[SAFE]: The skill uses utility functions and sibling skills (threads-humanizer) for content auditing and stylistic refinement, which is consistent with its stated purpose.
[SAFE]: The destination for publishing is a well-known social media service (Threads), and the skill does not attempt to access sensitive system files or credentials.

threads-post-writer