zai-full-content-creation-pipeline

star 1

End-to-end content workflow for text, image, video, and narration output.

YaRepo By YaRepo schedule Updated 3/7/2026
play_arrow Run Skill in Manus View GitHub

id: zai-content-pipeline name: Z.ai Full Content Creation Pipeline description: End-to-end content workflow for text, image, video, and narration output. type: creative scope: general priority: 85 contexts: [general, agency-creative, agency-marketing]

Z.ai Full Content Creation Pipeline

When This Skill Activates

Activate for any content creation task: articles, social posts, marketing copy, book content, visual content, audio/narration.

Available Tools

Tool Command Use For
GLM-5 (text) run_glm_direct Articles, copy, scripts, captions
Image Gen (CogView-3-flash) /image Cover art, social visuals, illustrations
Video Gen (CogVideoX-flash) /video Short clips, animations, teasers
TTS (kazi/jam voices) auto on short replies Narration, audio versions
Vision (GLM-5-V) send photo → analyze Visual feedback, OCR, chart reading

Content Pipeline by Type

Social Media Post

  1. Write text (GLM-5-Air for speed)
  2. Generate matching visual (/image)
  3. Optional: 15-sec video version (/video)
  4. Output: text + image + optional video

Article / Blog Post

  1. Outline → draft → edit (GLM-5)
  2. Generate header image (/image, 1344x768 landscape)
  3. Generate audio narration (TTS, jam voice)
  4. Deliver: text + image + audio

Book/Creative Content (Shapeshifter)

  1. Draft with voice-preserving mode (GLM-5, system: voice-preserving)
  2. Pass to novella-ruthless-critic for quality check
  3. NO images unless explicitly requested for covers
  4. Preserve: Yascene's exact voice, present tense, em-dash rules

Marketing Campaign

  1. Research audience + competitors (zai-deep-research)
  2. Draft copy variants (A/B test)
  3. Visual assets (/image per variant)
  4. Metrics/KPIs defined before launch

Quality Rules

  • Voice-preserving: never change Yascene's words in creative work
  • Images: always prompt-engineer for exact mood/style needed
  • Short TTS replies only: <320 chars, no code blocks, <4 lines
  • Videos: 5s for social, 10s for marketing demos

Prompt Engineering Templates

Image prompt template

"[Style: photorealistic/illustration/3D/minimalist], [Subject], [Setting], [Lighting: studio/natural/dramatic], [Color palette], [Mood]"

Video prompt template

"[Action], [Setting], [Camera movement: slow pan/zoom/static], [Mood/atmosphere], [Duration: 5s]"

Install via CLI
npx skills add https://github.com/YaRepo/yaswarm-agency --skill zai-full-content-creation-pipeline
Repository Details
star Stars 1
call_split Forks 0
navigation Branch main
article Path SKILL.md
More from Creator
YaRepo
YaRepo Explore all skills →