Use Merit Fit
Building
Changelog
Detail pages, funding tables, and library layout
Use Merit Fit got a long pass over its detail pages and funding sections. Questions, answers, reference material, and notes were rearranged into tighter, more consistent layouts, with several labels, descriptions, and spacing rules corrected along the way. The library and dashboard also picked up finishing work, including regrouped navigation and clearer section structure. Tables were aligned and paged on the client side, and a set of internal tests was fixed after the surrounding changes.
Safer extraction and better application review
This week focused on making extraction and comparison behave more safely around user-supplied information. Re-extraction now avoids overwriting what was already entered, corrected facts stay fixed, and the compare flow was repaired after issues with ids and test coverage surfaced. The work also added clearer progress reporting and removed cases where adding information could accidentally erase existing data. Application handling was tightened as well. Imports can bring in a whole shortlist at once, deadlines are shown with urgency in a single sortable list, and supporting details such as links, screenshots, notes, and version history were reorganized so they are easier to keep alongside an application. A set of testing-driven fixes also landed around stages, paste limits, URL input, and the master’s and doctoral scope.
Import review and admin cleanup updates
Use Merit Fit now shows the overview with real counts and replaces raw enum values in selects with readable labels. The wordmark also links back to home from every header. The import and admin flows were tightened up around rerouted URLs, purge counts, and screenshot handling. Imported rows can now be deleted one at a time, screenshots can be added during import without replacing existing ones, and the destructive-action pattern was documented along with the remaining open items.
Recover blocked pages and add account deletion
This week focused on making extraction, setup, and recovery clearer. Blocked pages can now be reopened with their source shown, rejected pages are treated as temporary readings rather than permanent dead ends, and the landing page and dashboard were adjusted to surface the product and setup earlier. Several import and form flows were also corrected, including pasted text handling, tab routing, and field prompts for study and funding details. There was also internal work around Neon usage and worker behavior. The product now measures its own duty cycle from exposed usage data, idle Chromium is released, and the boot and autosuspend assumptions were corrected. Account deletion was added with care for other users' data, and broken or fragile cases were written down so they can be tracked.
Library navigation, outcomes, and application notes
The Library now appears in navigation, and its kinds are organized as tabs. Application work also expanded with outcomes, a readiness timeline, notes, and links back and forth from the application side. Several interaction fixes landed across the tracker and sidebar, including clearer document titles, a visible SOP location, and a calmer attachment flow that now uses a dialog. Rewrapping text no longer counts as a page change, and the application view was tightened so only one primary action is presented at a time.
Programs, applications, and journey updates
This week added a new programs domain with data models, API routes, extraction, moderation, notifications, and a full web surface for browsing, saving, importing, comparing, and administering programs. The extraction flow was extended to keep more page content, crawl sub-pages, and capture deadline kinds, accepting-applications status, curriculum details, and related scholarship links. Notifications were deduplicated, deadline reminders were added, and the public copy and legal text were updated to match the shipped product. The user journey was also reshaped around applications and match gaps. The old tables were folded into an Application view with document attachment and score-report parsing, while the navigation, dashboard, profile, and workspace were reorganized around the new structure. Several backlog items were recorded as product decisions, and the week also included internal cleanup such as loading states, test adjustments, and removal of an unused web-side client.
Public site pages, SEO, and navigation
The public site was expanded with privacy, terms, contact, and learn pages, and the learning section was surfaced in navigation. The home page copy was tightened to better reflect the actual flow through the product. Search and sharing support was added with robots, sitemap, metadata, brand structured data, favicon and touch icons, and social preview images. Analytics tracking was gated behind consent, app error and not-found boundaries were added, and a few operational fixes landed, including cron access handling and source page repointing.
Split into API and frontend with shared auth
Use Merit Fit moved through a major architecture split this week. The app was separated into a frontend and an API service, with JWT-based sessions, shared cookie auth, and the main product areas running through the API. Deployment and environment setup were updated for the new three-service topology, including docs for local development, cron endpoints, and shared domain cookies. Several product areas were also updated alongside the split: the scholarship library, profile, writing workspace, dashboard advisor, feedback board, moderation, notifications, resume import, and admin flows all saw schema or application changes. The profile and writing experiences were redesigned around document vaults, autosave, versioned answers, and AI drafting, while scholarship import and deduplication gained tests and the storage layer moved user documents out of the database and into S3-compatible storage.