How konlu/story-studio works
Story Studio is an AI-agent skill for creating historical and biographical short-video assets, from topic and script through voice, timed visuals, and an editable Jianying draft. The README documents both a full workflow and modular handoffs; image generation and semantic review are agent-guided stages, not fully represented by sampled runtime code. The graph preserves those capabilities while keeping optional local rendering separate. Edges follow README workflow statements or visible imports/calls; unsampled subsystems use existing directory paths.
Main entry points: SKILL.md Agent Skill · doubao_tts.py Doubao TTS · export_jianying.py Jianying Export · align_with_coli.py Shot Alignment
Components 19
Story and Script
- Agent Skill
SKILL.md - Character Writing
modules/characterstory-writing/GUIDE.md - Historical Writing
modules/history-writing/scripts/history_tool.py - Story Corpus
modules/characterstory-writing/corpus - History Corpus
modules/history-writing/corpus
Voice and Visuals
- Voice Settings
scripts/voice_settings.py - Doubao TTS
scripts/doubao_tts.py - Voice Audio
- Coli Transcription
modules/history-image/scripts/transcribe_with_coli.py - Storyboard and Images
modules/history-image - Style References
modules/history-image/references
Timing and Delivery
- Shot Alignment
modules/render-history-video/scripts/align_with_coli.py - Review Checks
scripts/check_handoff.py - Motion Planning
modules/render-history-video/scripts/motion.py - Jianying Export
modules/render-history-video/scripts/export_jianying.py - Local Video Render
modules/render-history-video/scripts/render.py
Runtime Support
- Runtime Discovery
scripts/studio_runtime.py
Other components
- Creator
- Doubao Speech
Connections 21
- Creator → Agent Skill: requests creation
README.md - Agent Skill → Character Writing: guides writing
README.md - Agent Skill → Historical Writing: guides writing
README.md - Character Writing → Story Corpus: uses examples
README.md - Historical Writing → History Corpus: searches corpus
README.md - Agent Skill → Voice Settings: configures voice
README.md - Voice Settings → Doubao TTS: provides profile
scripts/doubao_tts.py - Doubao TTS → Doubao Speech: synthesizes with
README.md - Doubao TTS → Voice Audio: produces audio
README.md - Creator → Voice Audio: supplies recording
README.md - Voice Audio → Coli Transcription: transcribes
modules/history-image/scripts/transcribe_with_coli.py - Agent Skill → Storyboard and Images: plans shots
README.md - Storyboard and Images → Style References: anchors style
README.md - Voice Audio → Shot Alignment: times cuts
README.md - Coli Transcription → Shot Alignment: provides timestamps
- Shot Alignment → Review Checks: supports review
- Review Checks → Jianying Export: checks delivery
scripts/check_handoff.py - Jianying Export → Motion Planning: uses motion logic
modules/render-history-video/scripts/export_jianying.py - Jianying Export → Local Video Render: uses render helpers
modules/render-history-video/scripts/export_jianying.py - Agent Skill → Local Video Render: offers local render
README.md - Doubao TTS → Runtime Discovery: discovers tools
scripts/doubao_tts.py