Tag created by softprops/action-gh-release is the single source of truth.
Version bump in tauri.conf.json + Cargo.toml is local to the build only.
Co-Authored-By: Claude <noreply@anthropic.com>
- Compute version from latest git tag (bump patch)
- Update tauri.conf.json and Cargo.toml with new version
- Create GitHub Release with signed APK + AAB via softprops/action-gh-release
- Release name: v{version}, APK filename: zeavis-edu-v{version}.apk
Co-Authored-By: Claude <noreply@anthropic.com>
- Fix reference error in `DiagnosisResultView` by correctly mapping `riskLevel` prop.
- Integrate dynamic `RiskBadge` with automated severity-to-color mapping for both diagnosis and catalog views.
- Update icons in the medicine recommendation section of `catalog-page` and `DiagnosisResultView` for better visual consistency and intuition.
- Trigger on push to main (web/tauri/shared changes), PR, and manual
- Builds Vite frontend + Tauri Rust for 4 Android arches
- Signs APK with keystore from GitHub secrets
- Uploads signed APK + AAB as artifact
Co-Authored-By: Claude <noreply@anthropic.com>
- Replace previous icons in the medicine recommendation section with more intuitive visual elements.
- Enhance the user interface consistency within the catalog page.
- Implement `useEffect` to ensure smooth scroll-to-top on navigation.
- Add fallback logic to populate medicine recommendations from local seed data if backend payload is incomplete.
- Integrate dynamic `RiskBadge` component with automated severity-to-color mapping for better visual clarity.
- Import diseaseCatalogSeed into diagnosis detail and scan pages.
- Implement fallback logic to populate medicine recommendations when backend payload lacks treatment data.
- Ensure consistent educational content display across catalog and diagnosis views.
- Remove `onClick` toggle handler from `Sidebar` links to prevent unwanted state changes during navigation.
- Remove aggressive `onClick` auto-collapse from `MainLayout` content wrapper so the sidebar remains stable.
- Ensure sidebar state is exclusively managed by the dedicated toggle button, aligning with standard desktop UX.
- Restore setFileName(null) and setPreviewUrl(null) in the mutation onSuccess callback.
- Ensure the file upload area resets to its default state immediately after a successful scan, preparing the UI for the next input.
- Update and clarify inline code comments in the diagnosis result view component.
- Improve code readability and maintainability without altering any underlying logic or UI behavior.
- Remove nested overflow and fixed height classes (`max-h-[75vh]`, `overflow-y-auto`) from the inner modal container to fix the double scrollbar UI glitch.
- Adjust the Modal `size` property to a supported type to resolve TypeScript compilation errors.
- Ensure a cleaner, more seamless scrolling experience for the diagnosis result pop-up.
- Move the `onClick` event handler from the accordion header to the outermost card container in `catalog-page.tsx`.
- Allow users to click anywhere on the expanded card (including the description and treatment sections) to collapse it.
- Improve overall user experience by providing a significantly larger tap target, especially beneficial for mobile or touch-screen users.
- Import `diseaseCatalogSeed` to use as a fallback data source in `catalog-page.tsx`.
- Implement logic to merge backend API responses with local seed data, ensuring `imageUrl` and `medicineRecommendations` are populated.
- Fix an issue where disease images and specific treatments were not rendering due to missing fields in the database payload.
- Enhance UI resilience by displaying accurate visual and medical data without requiring immediate backend updates.
- Convert static PNG image assets to WebP format and remove old PNG files to significantly reduce image payload size and improve page load performance.
- Enhance overall UI resilience and frontend speed.
- calibrate_probs_selects_top_label: use logits that produce confidence >= 0.70
- prediction_response: forward status from Prediction struct instead of hardcode
Co-Authored-By: Claude <noreply@anthropic.com>
- PredictionResponse::status was missing in prediction_response()
- Prefix preprocess_start with underscore for unused variable warning
Co-Authored-By: Claude <noreply@anthropic.com>
Files on HF hub sit at repo root (model.onnx, etc.),
not in model/ subfolder. Fix download script to copy
each file to the correct subdirectory expected by Docker.
Co-Authored-By: Claude <noreply@anthropic.com>
- Integrate Convolutional Block Attention Module (CBAM) for improved feature focus
- Implement temperature scaling and confidence-based status reporting
- Automate dataset acquisition using kagglehub
- Update ONNX opset to 18 and refine preprocessing validation
- Import `useUiStore` into `main-layout.tsx` to access the global sidebar state.
- Add an `onClick` event listener to the main content wrapper.
- Automatically close the sidebar if it is expanded and the user clicks outside of it.
- Enhance overall UX by providing a more intuitive, frictionless way to dismiss the navigation menu.
- Update the `onClick` handler for navigation links in `sidebar.tsx` to use `!isSidebarOpen` instead of a hardcoded boolean.
- Enable navigation items to act as smart toggles: expanding the sidebar when clicked in collapsed mode, and collapsing it when clicked in expanded mode.
- Improve overall user experience by reducing the need to manually click the main toggle button.
- Move `isSidebarOpen` state from local component (`useState`) to global Zustand store (`ui-store.ts`).
- Fix an issue where the sidebar would incorrectly reset to its default closed state when navigating between routes (e.g., to the Telemetry page) due to component remounting.
- Ensure consistent sidebar toggle behavior across the entire dashboard application.
- Update `isSidebarOpen` initial state from true to false in `sidebar.tsx`.
- Ensure the sidebar starts in a mini/collapsed mode when the dashboard is first loaded to provide more space for the main content.
- No need to install TensorFlow (3GB) — just pip install huggingface_hub
- Download pre-built model.onnx directly (saved from notebook export)
- Cuts CI time from ~10 min to ~30 seconds
Co-Authored-By: Claude <noreply@anthropic.com>
- Verify HF_TOKEN exists before attempting download
- Group CI log output for readability
- Use os.environ['HF_TOKEN'] instead of .get() to fail fast
Co-Authored-By: Claude <noreply@anthropic.com>
- Remove binary model artifacts tracked in repo (.tflite, saved_model, tfjs model.json)
- CI now downloads best_model.keras from Hugging Face Hub (MythEclipse2737/corn-leaf-disease-classifier)
- Add save_model.py step before convert_onnx.py in CI pipeline
- Remove LFS patterns from .gitattributes (no longer needed)
- Remove lfs:true from checkout action
The model is fetched at CI time using HF_TOKEN secret — no large
binary files stored in git.
Co-Authored-By: Claude <noreply@anthropic.com>
