Update Log
Stay up-to-date with the latest features, improvements, and fixes for domyassessment.ai.
Introduced significant user experience improvements including an editable AI plan, citation style selection, robust PDF extraction, and drag-and-drop uploads. Overhauled the input UI for greater clarity and ease of use.
Key Changes:
- Added an editable textarea for the 'Assessment Plan' (Step 1 output), allowing users to review and modify the AI's plan before generating the final assessment.
- Implemented a dropdown menu for users to explicitly select a citation style (APA, MLA, etc.), providing direct control over referencing.
- Significantly improved PDF text extraction with a multi-layered fallback system, including OCR, to handle complex or image-based PDFs.
- Added drag-and-drop functionality to all file inputs, streamlining the upload process.
- Implemented contextual help tooltips on all input sections to provide clear, on-demand guidance.
- Added a 'Reset' button to revert any user edits made to the AI's assessment plan.
- Redesigned the file input UI to remove the browser's default 'No file chosen' text and resolve visual overlap bugs.
- Refined the UI for the multi-file upload list for better readability.
Improved AI's understanding of 'Additional Material' and 'Past Work' for more relevant outputs and better writing style emulation. Enabled multi-file uploads for 'Additional Material' and 'Past Work'. Refined UI spacing and added a new site favicon.
Key Changes:
- Enhanced the AI's system prompt to deeply analyze 'Additional Material', identify citable sources within them, and make this content central to its assessment planning. This aims to produce more contextually relevant and well-sourced outputs.
- Improved AI's system understanding of 'Past Work' to note and apply the users writting style, minor writting quirks, etc which greatly makes the final output sound not just more human, but as if the user wrote it themselves.
- Implemented multi-file upload capability for the 'Additional Material' section and 'Past Work', allowing users to provide comprehensive context from various sources concurrently, and to expand AI's understanding of the user's writing style.
- Reduced spacing between input sections (Task Sheet, Additional Material, Rubric, etc.) on the main assessment form for a more compact user interface.
- New favicon
Introduced new Info and Help pages, this update log, and made subtle improvements to header UI and navigation. Also updated AI Agent 2 prompt and addressed some spacing.
Key Changes:
- Added 'Info' button and `/info` page with detailed guidance on using the tool effectively.
- Added 'Help' button and `/help` page with contact information and links to the Info page.
- Added 'v0.7' button linking to this Update Log page in main and sticky headers.
- Increased 'Scroll to Start' text size for better visibility.
- Configured sticky navigation bar to appear sooner when scrolling down.
- Reduced bottom margin of the main header section to decrease space before the upload section.
- Updated AI Agent 2 prompt to re-emphasize understanding and applying content from 'Additional Material'.
- Addressed minor spacing issues within the upload sections for a tighter layout (prior to v0.8's major spacing overhaul).
Replaced Lenis smooth scrolling with native browser smooth scrolling for broader compatibility and performance.
Key Changes:
- Removed Lenis library and related context/provider components.
- Updated scroll-to-element logic in `app/page.tsx` to use `window.scrollTo`.
- Ensured `html { scroll-behavior: smooth; }` is active in `globals.css`.
Implemented an admin dashboard to monitor content uniqueness and system statistics. Enhanced the uniqueness service with Redis integration and fallback mechanisms.
Key Changes:
- Created `/admin/dashboard` page with key metrics.
- Developed `UniquenessService` for tracking content similarity and topic rotation.
- Integrated Redis for persistent storage with in-memory fallback.
- Added `ads.txt` for AdSense.
Enabled file uploads via Vercel Blob storage and implemented server-side logic for extracting text from PDF and DOCX files. Implemented client-side image compression.
Key Changes:
- Added `/api/upload-blob` route for handling file uploads.
- Integrated `unpdf`, `pdf-parse` for PDF text extraction with fallbacks.
- Used `mammoth` for DOCX text extraction.
- Implemented `InputSwitcher` component for choosing between file upload and text paste.
- Added client-side image compression using `browser-image-compression` for all image uploads to improve speed and reduce storage.
Developed the two-agent AI system for generating assessment plans and final assessments using OpenAI.
Key Changes:
- Created `app/actions.ts` with `generateAssessmentPlan` (Agent 1) and `generateFinalAssessment` (Agent 2).
- Implemented logic for handling various input types (task sheet, rubric, etc.).
- Designed prompts for AI agents to ensure quality and uniqueness.
Established the basic UI structure using Next.js, Tailwind CSS, and shadcn/ui components. Implemented the main page layout and form components.
Key Changes:
- Set up `app/page.tsx` with the main assessment form.
- Styled components using Tailwind CSS and custom global styles.
- Integrated shadcn/ui components like Card, Button, Input, Textarea.
- Added Progress Bar component.
Initial project setup with Next.js and basic configuration.
Key Changes:
- Initialized Next.js project.
- Configured Tailwind CSS and PostCSS.
- Set up basic layout and global styles.
- Added favicon and metadata.