InkRelay: Author-Led AI Content Production

InkRelay is an open-source content-production Skill I designed around an author’s materials, voice, and publishing goals. It coordinates an approved core text and optional visuals rather than offering one-click writing or a fixed renderer.

My contributions

• Long- and short-form workflows: Sources and argument structure precede section-level drafting; short content retains one core draft across optional formats.
• Author approval points: Distinguish drafts, approved text, and derivative assets without treating experiments as consent to publish.
• Cover-generation guidance: Translate author identity, the article’s argument, a single visual concept, and approved examples into prompts.
• Platform-aware production: Plan horizontal covers, optional square crops, 3:4 cards, and 9:16 text videos without requiring every format.
• Tool and asset boundaries: Orchestrate available, authorized tools without embedding personal branding, models, fonts, music, or credentials.
• Iterative review: Separate content checks from visual inspection; retain revisions and scenario walkthroughs without presenting them as independent model evaluations.

Workflow diagram

Scope

The public project supplies a Skill and production guidance, not a bundled renderer. Actual media requires available tools and authorization; paid generation, posting, and scheduling are not automatic.

GitHub source and documentation · Skill entry