@waterwx/dsh-novel-forge 1.7.2 → 1.7.3

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (25) hide show
  1. package/README.md +1 -1
  2. package/lib/client.js +918 -287
  3. package/lib/client.js.map +1 -1
  4. package/lib/index.js +372 -42
  5. package/lib/index.js.map +1 -1
  6. package/lib/types/protocol.d.ts +14 -10
  7. package/lib/types/shot-language.d.ts +46 -0
  8. package/lib/types/story-beat-language.d.ts +35 -0
  9. package/package.json +4 -2
  10. package/skills/chapter-batch/SKILL.md +81 -0
  11. package/src/client/panel/CreateBookView.tsx +4 -0
  12. package/src/client/panel/ImportModal.tsx +11 -2
  13. package/src/client/panel/NovelPanel.tsx +1 -0
  14. package/src/client/panel/RunPanel.tsx +1 -1
  15. package/src/client/panel/SceneLibrary.tsx +1 -1
  16. package/src/client/panel/ShelfView.tsx +2 -2
  17. package/src/client/panel/StoryboardTab.tsx +42 -10
  18. package/src/engine.ts +35 -14
  19. package/src/index.ts +1 -0
  20. package/src/protocol.ts +22 -10
  21. package/src/routes.ts +13 -7
  22. package/src/run.ts +5 -4
  23. package/src/shot-language.ts +181 -0
  24. package/src/story-beat-language.ts +130 -0
  25. package/lib/types/client/panel/RoleVisualPanel.d.ts +0 -15
package/README.md CHANGED
@@ -169,7 +169,7 @@ tsdown.config.ts 双面打包配置 / dual-face bundling config
169
169
  - 生成消耗 LLM API 额度（默认 deepseek-official / deepseek-v4-flash）/ generation consumes LLM API quota
170
170
  - 章节质量取决于大纲完整度 / chapter quality depends on outline completeness
171
171
  - 设置写入 `~/.dsh/settings.yaml` 的 `dsh-novel-forge` 段 / settings persist in `~/.dsh/settings.yaml`
172
- - 角色/场景视觉资产为「提示词 + 图集」形态，不内置生图能力；出图请使用外部工具（即梦/豆包/Krea 等），提示词可直接复制 / visual assets are prompt+image anchors; image generation is external (Jimeng/Doubao/Krea) with copy-ready prompts
172
+ - 角色/场景视觉资产为「提示词 + 图集」形态，并可在设置页配置内置生图模型（imageModels/imageApiKey，支持多模型库与连通性测试）；出图既可走内置生成，也可用外部工具（即梦/豆包/Krea 等），提示词可直接复制 / visual assets are prompt+image anchors, with optional built-in image-model configuration (imageModels/imageApiKey, multi-model library + endpoint test); images may be generated in-app or via external tools (Jimeng/Doubao/Krea) with copy-ready prompts
173
173
  - 漫剧分镜/漫画脚本功能已移除（v1.3.0），角色形象与场景库保留为漫剧流水线的视觉锚点底座 / storyboard & comic-script features removed in v1.3.0; role visuals & scene library remain as the visual-anchor foundation
174
174
 
175
175
  ---