ddCat
a3bf0bd801
feat(canvas): 多模态表格 - 合并 PR #549 并保留本地交互与校验
2026-09-20 17:18:46 +08:00
Codex
6d355380cb
feat(canvas): support multimodal video storyboard tables
2026-09-20 12:59:23 +08:00
Codex
0e61fb00ab
fix: honor requested list rows and preserve ai table data
2026-09-19 20:51:55 +08:00
Codex
3f5b0bd347
fix: route list mode through multimodal text model
2026-09-19 20:04:26 +08:00
Codex
af96824011
fix: harden list mode and batch generation flow
2026-09-19 15:35:11 +08:00
Codex
944a220939
fix: force text model for list mode to avoid image model rejection
2026-09-19 14:46:26 +08:00
Codex
937f2b7a3a
fix: use buildGenerationConfig for list mode to resolve channel/model properly
2026-09-19 14:39:09 +08:00
Codex
e921deac74
feat: add list mode for AI-powered batch table generation
...
- Add text/list toggle in prompt panel
- List mode: AI analyzes connected images + prompt, outputs structured JSON table
- Batch table node renders AI-generated text columns as read-only content
- Column names and row count determined by AI (3-6 columns, N rows)
2026-09-19 14:25:24 +08:00