| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
feat(autolab): Phase 5 — tests + AI streaming + code extraction Phase 5.1: Backend + Frontend unit tests - Add 11 Rust tests for NotebookActor (session lifecycle, execution, dirty upstream re-execution, diagnostics extraction) - Add 14 Vitest tests for useNotebook (cell CRUD, move, dirty tracking, AutoDown serialization/deserialization, code extraction) Phase 5.2: AI streaming via SSE - Backend: POST /api/notebook/{sid}/ai-stream returns Sse<Event> - ClaudeProvider::stream_chat() pipes reqwest chunks through mpsc - Frontend: askAIStream() using EventSource with onMessage callback - AIChatBar.vue updated to use streaming API Phase 5.3: One-click code extraction - extractCodeFromAI(cellId) parses fenced code blocks from AI response - CellToolbar.vue adds 'Extract code' button (Code2 icon) for AI cells - App.vue wires extract-code event across CellCanvas/CellItem/CellToolbar - Alerts user if no code block found Also fixes: - loadFromAd trailing newline no longer creates empty cell - isDirty exposed in useNotebook return object - table type included in CellType list Refs: Plan 246 | 5 个月前 | |
feat(autolab): Phase 2 & 3 — frontend skeleton, cell types, AI integration, dependency re-execution Phase 2: AutoLab Frontend Skeleton - Scaffold packages/auto-lab-ui (Vite + Vue 3 + TypeScript) - Implement CellCanvas + CellItem with add/delete/reorder/collapse/status - Implement AIChatBar (bottom input → cell stream) - Implement VariableInspector side panel - Implement .ad notebook file read/write (load/save/serialize) - Integrate auto-playground-vue CodeEditor component Phase 3: Cell Type System + AI Integration - Add Chart and Table built-in cell types with dedicated renderers - OutputChart.vue: SVG-based bar/line chart from JSON result - OutputTable.vue: HTML table from JSON array - Integrate Claude API backend (ClaudeProvider in notebook/ai.rs) - POST /api/notebook/{sid}/ai endpoint - System prompt with Auto language syntax rules - Frontend askAI() with preceding code cell context assembly - Dependency-chain incremental re-execution - Backend dirty cell tracking via cell_snapshots - Cascade dirty marking through depends_on graph - Automatic upstream re-execution before target cell Plan: docs/plans/246-autolab-design.md updated | 5 个月前 | |
feat(autolab): Phase 2 & 3 — frontend skeleton, cell types, AI integration, dependency re-execution Phase 2: AutoLab Frontend Skeleton - Scaffold packages/auto-lab-ui (Vite + Vue 3 + TypeScript) - Implement CellCanvas + CellItem with add/delete/reorder/collapse/status - Implement AIChatBar (bottom input → cell stream) - Implement VariableInspector side panel - Implement .ad notebook file read/write (load/save/serialize) - Integrate auto-playground-vue CodeEditor component Phase 3: Cell Type System + AI Integration - Add Chart and Table built-in cell types with dedicated renderers - OutputChart.vue: SVG-based bar/line chart from JSON result - OutputTable.vue: HTML table from JSON array - Integrate Claude API backend (ClaudeProvider in notebook/ai.rs) - POST /api/notebook/{sid}/ai endpoint - System prompt with Auto language syntax rules - Frontend askAI() with preceding code cell context assembly - Dependency-chain incremental re-execution - Backend dirty cell tracking via cell_snapshots - Cascade dirty marking through depends_on graph - Automatic upstream re-execution before target cell Plan: docs/plans/246-autolab-design.md updated | 5 个月前 | |
feat(autolab): Phase 5 — tests + AI streaming + code extraction Phase 5.1: Backend + Frontend unit tests - Add 11 Rust tests for NotebookActor (session lifecycle, execution, dirty upstream re-execution, diagnostics extraction) - Add 14 Vitest tests for useNotebook (cell CRUD, move, dirty tracking, AutoDown serialization/deserialization, code extraction) Phase 5.2: AI streaming via SSE - Backend: POST /api/notebook/{sid}/ai-stream returns Sse<Event> - ClaudeProvider::stream_chat() pipes reqwest chunks through mpsc - Frontend: askAIStream() using EventSource with onMessage callback - AIChatBar.vue updated to use streaming API Phase 5.3: One-click code extraction - extractCodeFromAI(cellId) parses fenced code blocks from AI response - CellToolbar.vue adds 'Extract code' button (Code2 icon) for AI cells - App.vue wires extract-code event across CellCanvas/CellItem/CellToolbar - Alerts user if no code block found Also fixes: - loadFromAd trailing newline no longer creates empty cell - isDirty exposed in useNotebook return object - table type included in CellType list Refs: Plan 246 | 5 个月前 | |
feat(autolab): Phase 5 — tests + AI streaming + code extraction Phase 5.1: Backend + Frontend unit tests - Add 11 Rust tests for NotebookActor (session lifecycle, execution, dirty upstream re-execution, diagnostics extraction) - Add 14 Vitest tests for useNotebook (cell CRUD, move, dirty tracking, AutoDown serialization/deserialization, code extraction) Phase 5.2: AI streaming via SSE - Backend: POST /api/notebook/{sid}/ai-stream returns Sse<Event> - ClaudeProvider::stream_chat() pipes reqwest chunks through mpsc - Frontend: askAIStream() using EventSource with onMessage callback - AIChatBar.vue updated to use streaming API Phase 5.3: One-click code extraction - extractCodeFromAI(cellId) parses fenced code blocks from AI response - CellToolbar.vue adds 'Extract code' button (Code2 icon) for AI cells - App.vue wires extract-code event across CellCanvas/CellItem/CellToolbar - Alerts user if no code block found Also fixes: - loadFromAd trailing newline no longer creates empty cell - isDirty exposed in useNotebook return object - table type included in CellType list Refs: Plan 246 | 5 个月前 | |
feat(autolab): Phase 2 & 3 — frontend skeleton, cell types, AI integration, dependency re-execution Phase 2: AutoLab Frontend Skeleton - Scaffold packages/auto-lab-ui (Vite + Vue 3 + TypeScript) - Implement CellCanvas + CellItem with add/delete/reorder/collapse/status - Implement AIChatBar (bottom input → cell stream) - Implement VariableInspector side panel - Implement .ad notebook file read/write (load/save/serialize) - Integrate auto-playground-vue CodeEditor component Phase 3: Cell Type System + AI Integration - Add Chart and Table built-in cell types with dedicated renderers - OutputChart.vue: SVG-based bar/line chart from JSON result - OutputTable.vue: HTML table from JSON array - Integrate Claude API backend (ClaudeProvider in notebook/ai.rs) - POST /api/notebook/{sid}/ai endpoint - System prompt with Auto language syntax rules - Frontend askAI() with preceding code cell context assembly - Dependency-chain incremental re-execution - Backend dirty cell tracking via cell_snapshots - Cascade dirty marking through depends_on graph - Automatic upstream re-execution before target cell Plan: docs/plans/246-autolab-design.md updated | 5 个月前 | |
feat(autolab): Phase 2 & 3 — frontend skeleton, cell types, AI integration, dependency re-execution Phase 2: AutoLab Frontend Skeleton - Scaffold packages/auto-lab-ui (Vite + Vue 3 + TypeScript) - Implement CellCanvas + CellItem with add/delete/reorder/collapse/status - Implement AIChatBar (bottom input → cell stream) - Implement VariableInspector side panel - Implement .ad notebook file read/write (load/save/serialize) - Integrate auto-playground-vue CodeEditor component Phase 3: Cell Type System + AI Integration - Add Chart and Table built-in cell types with dedicated renderers - OutputChart.vue: SVG-based bar/line chart from JSON result - OutputTable.vue: HTML table from JSON array - Integrate Claude API backend (ClaudeProvider in notebook/ai.rs) - POST /api/notebook/{sid}/ai endpoint - System prompt with Auto language syntax rules - Frontend askAI() with preceding code cell context assembly - Dependency-chain incremental re-execution - Backend dirty cell tracking via cell_snapshots - Cascade dirty marking through depends_on graph - Automatic upstream re-execution before target cell Plan: docs/plans/246-autolab-design.md updated | 5 个月前 | |
feat(autolab): Phase 2 & 3 — frontend skeleton, cell types, AI integration, dependency re-execution Phase 2: AutoLab Frontend Skeleton - Scaffold packages/auto-lab-ui (Vite + Vue 3 + TypeScript) - Implement CellCanvas + CellItem with add/delete/reorder/collapse/status - Implement AIChatBar (bottom input → cell stream) - Implement VariableInspector side panel - Implement .ad notebook file read/write (load/save/serialize) - Integrate auto-playground-vue CodeEditor component Phase 3: Cell Type System + AI Integration - Add Chart and Table built-in cell types with dedicated renderers - OutputChart.vue: SVG-based bar/line chart from JSON result - OutputTable.vue: HTML table from JSON array - Integrate Claude API backend (ClaudeProvider in notebook/ai.rs) - POST /api/notebook/{sid}/ai endpoint - System prompt with Auto language syntax rules - Frontend askAI() with preceding code cell context assembly - Dependency-chain incremental re-execution - Backend dirty cell tracking via cell_snapshots - Cascade dirty marking through depends_on graph - Automatic upstream re-execution before target cell Plan: docs/plans/246-autolab-design.md updated | 5 个月前 | |
feat(autolab): Phase 2 & 3 — frontend skeleton, cell types, AI integration, dependency re-execution Phase 2: AutoLab Frontend Skeleton - Scaffold packages/auto-lab-ui (Vite + Vue 3 + TypeScript) - Implement CellCanvas + CellItem with add/delete/reorder/collapse/status - Implement AIChatBar (bottom input → cell stream) - Implement VariableInspector side panel - Implement .ad notebook file read/write (load/save/serialize) - Integrate auto-playground-vue CodeEditor component Phase 3: Cell Type System + AI Integration - Add Chart and Table built-in cell types with dedicated renderers - OutputChart.vue: SVG-based bar/line chart from JSON result - OutputTable.vue: HTML table from JSON array - Integrate Claude API backend (ClaudeProvider in notebook/ai.rs) - POST /api/notebook/{sid}/ai endpoint - System prompt with Auto language syntax rules - Frontend askAI() with preceding code cell context assembly - Dependency-chain incremental re-execution - Backend dirty cell tracking via cell_snapshots - Cascade dirty marking through depends_on graph - Automatic upstream re-execution before target cell Plan: docs/plans/246-autolab-design.md updated | 5 个月前 | |
feat(autolab): Phase 5 — tests + AI streaming + code extraction Phase 5.1: Backend + Frontend unit tests - Add 11 Rust tests for NotebookActor (session lifecycle, execution, dirty upstream re-execution, diagnostics extraction) - Add 14 Vitest tests for useNotebook (cell CRUD, move, dirty tracking, AutoDown serialization/deserialization, code extraction) Phase 5.2: AI streaming via SSE - Backend: POST /api/notebook/{sid}/ai-stream returns Sse<Event> - ClaudeProvider::stream_chat() pipes reqwest chunks through mpsc - Frontend: askAIStream() using EventSource with onMessage callback - AIChatBar.vue updated to use streaming API Phase 5.3: One-click code extraction - extractCodeFromAI(cellId) parses fenced code blocks from AI response - CellToolbar.vue adds 'Extract code' button (Code2 icon) for AI cells - App.vue wires extract-code event across CellCanvas/CellItem/CellToolbar - Alerts user if no code block found Also fixes: - loadFromAd trailing newline no longer creates empty cell - isDirty exposed in useNotebook return object - table type included in CellType list Refs: Plan 246 | 5 个月前 |
| 文件 | 最后提交记录 | 最后更新时间 |
|---|---|---|
| 5 个月前 | ||
| 5 个月前 | ||
| 5 个月前 | ||
| 5 个月前 | ||
| 5 个月前 | ||
| 5 个月前 | ||
| 5 个月前 | ||
| 5 个月前 | ||
| 5 个月前 | ||
| 5 个月前 |