Sniper

Recruiter-facing browser for hackathon talent. Recruiters search for specific signals such as tech stack, hackathon, or project archetype, then move through ranked developers, project histories, podiums, and outgoing offer status.

The product is built around pre-LinkedIn signal: Devpost submissions are indexed and surfaced before a winner posts about the result elsewhere, giving recruiters a narrow outreach window.

No applicant-facing workflow. No OA flow. No scoring system.

Apps

backend/ is a Hono + TypeScript API on port 8080.

frontend/ is a Next.js 15 App Router UI on port 3000.

Deployment notes are in DEPLOYMENT.md. The free deployment path uses Vercel for both frontend/backend and GitHub Actions for scheduled scraping.

Run Locally

Backend:

cd backend
npm install
npm run dev

Frontend:

cd frontend
npm install
npm run dev

The frontend defaults to NEXT_PUBLIC_API_URL=http://localhost:8080.

Live Devpost Data

1. Run backend/supabase/schema.sql in the Supabase SQL editor.
2. Put SUPABASE_URL and SUPABASE_SERVICE_ROLE_KEY in backend/.env.
3. Scrape a Devpost gallery:

cd backend
npm run scrape:devpost -- --url https://thx.devpost.com/project-gallery --max-pages 1 --max-projects 24

Or automatically discover ended Devpost hackathons and scrape each project gallery:

npm run scrape:devpost:all -- --max-list-pages 1 --max-hackathons 5 --max-project-pages 1 --max-projects-per-hackathon 24

Automatic discovery prioritizes high-profile hackathons first using registration count, winner-announced state, recency, organizer names, and keywords.

Run the scraper continuously:

npm run scrape:devpost:watch -- --max-list-pages 3 --max-hackathons 10 --max-project-pages 1 --max-projects-per-hackathon 24

The watcher skips hackathons attempted in the last 24 hours by default, so it keeps moving through prioritized events instead of reprocessing the same few galleries every cycle.

For an uncapped run across all discovered ended hackathons:

npm run scrape:devpost:all -- --all --yes

Uncapped runs can take hours. Start with the capped command first.

4. Set DATA_SOURCE=supabase in backend/.env and restart the backend.

The same API routes then serve scraped Supabase data instead of the local seed.

Verification

cd backend && npm run build
cd frontend && npm run build

Core API

GET /api/developers

GET /api/developers/:id

GET /api/hackathons

GET /api/hackathons/:slug

GET /api/projects

GET /api/projects/:id

GET /api/feed/fresh

GET /api/graph

GET /api/offers

GET /api/admin/ingestion

POST /api/offers

PATCH /api/offers/:id