ReadEasy

Inspiration

Health literacy is a public-health barrier: per HHS Healthy People 2030, too many adults struggle with medical forms and benefits letters — and dense government sites make it worse. Walls of text, buried deadlines, and fixed small type lock out readers with dyslexia, ADHD, or low vision, and anyone facing a form on a deadline. Browser reader mode only strips styling; it never rewrites. I wanted a tool that opens that door: paste any health or benefits page and get back something a patient or caregiver can act on, written with person-first respect.

What it does

ReadEasy restructures any web page into clear, accessible formats. Paste a URL — or raw page text for blocked sites — and get a split view: the cleaned original on the left, the accessible version on the right. Modes include Focus, Dyslexia/Bionic, Action checklist (tasks with urgency and deadlines as Step 2 of 4), Listen karaoke, ADHD micro-cards, Ask this page, Reading level, plus cosmetic toggles, history, and export. Live demo: https://readeasy-gateway.vercel.app — code: https://github.com/GhostInHex/readeasy/tree/hackathon/gateway. Try the IRS, USCIS, and UT Dallas Apply demos to see benefits and enrollment text become checklists.

How I built it

One server route owns a four-stage pipeline: Fetch → Clean → Restructure → Render. Fetch grabs the HTML; Clean strips ads, nav, and scripts with Mozilla Readability plus jsdom (no AI). Restructure sends the cleaned text to OpenRouter with a strict JSON contract (title, summary, reading time, actionItems, sections) from lib/llm/prompt.ts, with person-first, plain-language rules. Render is client-side: every mode renders that JSON via a mode registry. The stack is Next.js 15, React, TypeScript, OpenRouter with fallbacks, browser speechSynthesis, and Vercel (no database, no auth).

Challenges I ran into

First, ssa.gov returns HTTP 403 to server fetches, so a raw-text paste path runs the identical restructure with no network. Second, the no-hallucination constraint: the prompt bans outside knowledge, copies names, dates, and amounts verbatim, and the original stays visible for verification. Third, demo reliability: cached fixtures (cleaned text plus screenshots) for the demo trio, schema-checked with verify:trio, so judging never depends on the network.

What I learned

Accessibility is restructuring, not restyling — grade-5 plain language plus extracted deadlines is what unblocks action. I learned to treat the LLM as a constrained rewriter with a validated schema, one retry, and structured errors. Most of all, social impact is measurable: the readability score (college level to grade 5), shown beside the original, lets any judge verify the delta in seconds.

What's next

Next is covering payer portals and after-visit summaries, adding multilingual variants, and piloting the checklist flow with a campus clinic. The goal stays the same: every reader finishes knowing what to do and by when.

Built With: Next.js 15, React, TypeScript, OpenRouter, Mozilla Readability, jsdom, Vercel, browser speechSynthesis.

Provenance

Built on the ReadEasy Buildverse base (Sep 1–6) and extended for GatewayHacks 2026 Track 1 Accessibility & Health (secondary: Equity in Education) with track framing, fixtures, and demo hardening.

Built With

• jsdom
• mozilla-readability
• next.js-15
• openrouter
• react
• typescript
• vercel