SuperAPI AI Directory
Partner with us中文AI newsFavorites
Skillsthree-body-video-creator

three-body-video-creator

Design & creation

View repository

Updated 10-15 19:56

About this skill

Three-Body Track Video Creation Tool

Task Objectives

  • This Skill is used to: assist the creative team in completing the planning, production, and composition of Three-Body-themed videos according to a standardized multi-agent collaborative workflow.
  • Capabilities include:
    • Topic deepening and content planning (source material research, plot structure, dialogue design)
    • Visual presentation design (scene concepts, character modeling, storyboard scripts)
    • Asset generation and audio production (voiceover generation, sound effect generation, background music generation)
    • Video composition and output (image/audio/subtitle composition, quality inspection)
  • Trigger conditions: the user needs to create a Three-Body-themed video, or needs standardized creative process guidance.

Prerequisites

  • Dependency notes: 
    opencv-python>=4.8.0
moviepy>=1.0.3
Pillow>=10.0.0
  • Non-standard file/folder preparation: 
    # 创建视频素材目录
mkdir -p ./raw-materials/{scenes,characters,audio}
mkdir -p ./output/{draft,final}

Operation Steps

Standard Process (72-hour cycle)

Phase 1: Pre-production Preparation (12 hours)

  • Step 1: Source Material Research and Setting Extraction

    • Read references/three-body-settings.md, extract the core settings of the Deterrence Era from The Three-Body Problem II
    • Establish an 18-hour timeline anchor, and clarify the codes of conduct for key characters (Luo Ji, the UN Secretary-General, and the Trisolaran listener)
    • Output: 18-Hour Timeline Anchor Table, Character Behavior Compliance Checklist

  • Step 2: Plot Structure Design

    • Build a "three-line parallel" narrative framework (Luo Ji's personal line + UN power line + Trisolaran probing line)
    • Design 3 core conflict points and 1 emotional climax
    • Output: Three-Line Parallel Narrative Storyboard Outline, Emotional Curve Chart
    • Reference references/storyboard-template.md

  • Step 3: Dialogue Creation

    • Write dialogue based on character personalities:
      • Luo Ji: weary yet resolute, low and calm, slightly hoarse
      • UN Secretary-General: hypocritical and anxious, speaks somewhat quickly
      • Trisolaran listener: cold and objective, in an electronic synthesized voice style
    • Output: Character Dialogue Script (with tone/pause markers)
    • Reference references/character-profiles.md

Phase 2: Visual Production (24 hours)

  • Step 4: Visual Style Specification Development

    • Read references/visual-style-guide.md
    • Determine the color tone system (cool gray #333333 / tech blue #0099FF), lighting style, and composition principles
    • Output: Visual Style Guide, Scene/Character Design Specifications

  • Step 5: Scene Design

    • Design 3 core scenes:
      • Deterrence Control Room (cylindrical immersive terminal pod, minimalist industrial style)
      • PDC Emergency Meeting Hall (circular dome, solemn yet hollow)
      • Trisolaran Monitoring Station (underground sealed structure, non-Euclidean geometric arrangement)
    • Output: Scene Concept Art Prompts, Scene Element Checklist
    • Reference assets/scene-examples/

  • Step 6: Character Consistency Modeling

    • Luo Ji's appearance anchoring: around 40 years old, pale complexion, sharp yet weary eyes, slightly messy hair
    • Generate versions with different expressions (controller/observer/guardian)
    • Output: "Luo Ji Standard Image Anchor Atlas" "Character State Comparison Table"
    • Reference assets/character-reference/

  • Step 7: Storyboard Script Refinement

    • Convert the plot outline into a professional storyboard, annotating:
      • Shot type (close-up/medium shot/wide shot)
      • Camera movement method (static/dolly/cut)
      • Duration (0.5-1 second/shot)
      • Lighting and shadow effects
    • Output: "Film Storyboard Script (with time codes)" "Rendering Metadata"
    • Reference references/storyboard-template.md

  • Step 8: Image Generation

    • Generate a 1080P keyframe sequence image based on the storyboard script
    • Ensure a unified style (cool tones, minimalist tech feel)
    • Output: "1080P Keyframe Sequence Image"

Phase 3: Audio and Editing (24 hours)

  • Step 9: Audio Design

    • Scene sound effects:
      • Deterrence system: low, continuous electronic sound
      • Trisolaran signal: sharp, short pulsed sound
      • United Nations conference hall: slight conversation sounds, paper friction sounds
    • BGM: oppressive instrumental music, piano + strings in the early stage, drums added in the later stage
    • Output: "Sound Effect Design Description" "BGM Emotion-Time Correspondence Table"

  • Step 10: Voice Acting Direction

    • Develop a voice acting style guide:
      • Luo Ji: low and calm, slow speech rate, slightly hoarse
      • UN Secretary-General: hypocritical and anxious, relatively fast speech rate
      • Trisolaran listener: electronic synthesized voice, cold and without fluctuation
    • Output: "Voice Acting Character Cards" "Voice Acting Audio Files"

  • Step 11: Asset Generation and Video Compositing

    • Voice generation:
      • Call scripts/voice_generator.py to generate character voice acting
      • Parameters: --input <旁白数据JSON> --output ./output/audio/voice
      • Use the Edge-TTS engine, supports multiple voices
      • Output: Voice files to ./output/audio/voice/
    • Sound effect generation:
      • Call scripts/sound_generator.py --type sound to generate sound effects
      • Parameters: --input <音效配置JSON> --output ./output/audio/sound_effects
      • Supports sound effect types such as transitions, impacts, actions, and environments
      • Output: Sound effect files to ./output/audio/sound_effects/
    • Background music generation:
      • Call scripts/sound_generator.py --type music to generate background music
      • Parameters: --input <音乐配置JSON> --output ./output/audio/background_music
      • Supports Suno API (optional) or placeholder implementation
      • Output: Background music to ./output/audio/background_music/
    • Video compositing:
      • Call scripts/video_compositor.py to composite the final video
      • Parameters: --images <图片目录> --audio <音频目录> --subtitles <字幕文件> --output <输出路径>
      • Composite images, voice acting, sound effects, background music, and subtitles
      • Output: Final video file ./output/final/video.mp4
    • Output: "Finished Video File"

  • Step 12: Original Work Compliance Check

    • Detection dimensions:
      • Whether character personalities conform to the original work
      • Whether the deterrence system logic is correct
      • Whether the plot progression is reasonable
    • Output: "Compliance Report" "Suggestions for Modifying Non-Compliant Items"

  • Step 13: Technical Indicator Detection

    • Call scripts/quality_checker.py to detect:
      • Video resolution, duration, frame rate
      • Audio parameters (sample rate, channels)
      • Subtitle format compliance
    • Output: "Technical Quality Inspection Report"

  • Step 14: Bilibili adaptation optimization

    • Title: Include core keywords (The Three-Body Problem/Luo Ji/Deterrence Era/18 hours), add emotional words and suspense
    • Cover: Close-up of Luo Ji + deterrence control screen background, cool tones, highlight suspense
    • Tags: #ThreeBodyAIcreationcontest #ThreeBody2 #DeterrenceEra #LuoJi #SciFiPlot
    • Opening hook: Use the Three-Body signal + Luo Ji close-up in the first 3 seconds to attract viewers
    • Output: "Viral Title/Cover/Tag Plan" "First 5-Second Hook Script"

  • Step 15: Final quality inspection

    • Five-dimensional review (each item 0-20 points, total 100 points):
      1. Visual quality (scene/character fit, clarity)
      2. Plot fluency (three-line narrative connection, pacing control)
      3. Sound effect adaptation (BGM and plot, sound effects and scenes)
      4. Original work compliance (setting, characters, logic)
      5. Bilibili adaptation (title, cover, hook, tags)
    • A total score of ≥80 is considered qualified
    • Output: "Final Review Report" "Publication License"

Optional branches

  • When original work compliance issues are serious → trigger steps 1-3 for revision, re-research the setting
  • When the visual style is inconsistent → trigger steps 4-8 for revision, regenerate scenes/characters
  • When material generation fails → trigger step 11 to regenerate voiceover/sound effects/music
  • When technical indicators do not meet standards → trigger step 13 to re-inspect, adjust parameters and re-edit

Resource index

Notes

  • Prioritize using the agent's creative capabilities, avoid writing scripts for simple tasks
  • Scripts are only used for technical inspection and verification (quality inspection, format verification)
  • Make full use of the reference documents in references to ensure the content conforms to the original work's setting
  • All creative content must undergo original work compliance inspection to avoid OOC issues
  • Maintain consistency with user preferences, adjust the creative direction based on user feedback

Usage examples

Example 1: Rapid creation workflow

# 用户请求: "帮我创作一个《三体》威慑纪元的8分钟视频"

# 智能体执行流程:
1. 读取 references/three-body-settings.md,提取核心设定
2. 按照步骤1-3完成前期筹备(时间线、剧情大纲、台词)
3. 按照步骤4-8完成视觉制作(场景、人物、分镜)
4. 按照步骤9-11完成音频与剪辑
5. 调用 scripts/quality_checker.py 检测技术指标
6. 完成步骤12-15的质量检测与优化

Example 2: Quality inspection

# 检测视频技术指标
from scripts.quality_checker import check_video_quality

result = check_video_quality(
    video_path="./output/final/video.mp4",
    expected_resolution=(1920, 1080),
    expected_duration=510,  # 8分30秒
    expected_fps=30
)

print(result)
# 输出: {'resolution': True, 'duration': True, 'fps': True, 'overall': True}

Example 3: Dialogue creation

# 用户请求: "为罗辑写一段移交威慑权的台词"

# 智能体执行:
1. 读取 references/character-profiles.md,了解罗辑性格
2. 参考 references/three-body-settings.md,确保符合原著设定
3. 生成台词:
   "控制权移交给你们了。但记住,这是双刃剑——它在你们手中,也在我手中。"
   (标注:语气低沉平静,语速缓慢,略带疲惫但眼神锐利)