Content Distiller

Problem context

Content creators such as book-club leaders, web-novel authors, course instructors, and domain experts often hold valuable materials: reading notes, novel text, course decks, expertise, and speaking style. These assets are useful, but they are usually delivered one time or through manual consultation. Content Distiller turns them into structured AI skills that readers, students, or clients can repeatedly question.

Core workflow

It organizes the process around four input scenarios: book reading, novel reading, course learning, and AI avatar. Each scenario expects different inputs. Books need title and reading notes; novels need text, copyright statement, and character style; courses need PPT, video, or Markdown materials; AI avatars need professional skills, speaking-style samples, and service list.

After the user provides materials, it generates 15-50 high-frequency Q&A items. The Q&A is adjusted into a GEO-friendly structure using FAQ + EEAT + data anchors and scored across depth, coverage, differentiation, compliance, and value. Publishing continues only when the score is above 80. Pricing and distribution are also required: free content goes to the matrix distribution assistant, while one-time, monthly, or yearly paid content goes to the paid publishing assistant.

Limits and notes

• Copyright: novel scenarios require a copyright statement.
• Multimodal: image generation and character presentation depend on WorkBuddy.
• Paid publishing: one-time, monthly, or yearly paid skills require platform membership.
• Completeness: the skill must include complete files and metadata such as name/slug/tags/category/language/price.
• Co-production: the flow automatically credits Qingmu Hui Jianghu.