KolQuest Docs

First step: fetch the raw HTML and identify the tech stack.

curl -s 'https://kolscan.io/leaderboard' -o page.html
head -100 page.html

Found /_next/static/chunks/ script tags — confirming it's a Next.js application with App Router.

Tried common REST patterns with GET requests:

curl -s -o /dev/null -w "%{http_code}" 'https://kolscan.io/api/leaderboard'  → 400
curl -s -o /dev/null -w "%{http_code}" 'https://kolscan.io/api/kols'          → 400
curl -s -o /dev/null -w "%{http_code}" 'https://api.kolscan.io/'              → 401

400 means the endpoints exist but reject GET requests. 401 on the API subdomain means auth required.

Extracted all script chunk URLs from the page source:

curl -s 'https://kolscan.io/leaderboard' \
  | grep -oE '"[^"]*/_next/static/chunks/[^"]*"'

Found key chunks including:

• app/leaderboard/page-939ad755c42d8b9d.js — the leaderboard page
• 184-3d6ce3be6906820b.js — shared API helper chunk

Searched each JS chunk for /api/ paths:

for chunk in 341-*.js 184-*.js 255-*.js; do
  curl -s "https://kolscan.io/_next/static/chunks/${chunk}" \
    | grep -oP '"/api/[^"]*"'
done

Chunk 184 contained the gold: /api/trades, /api/tokens, /api/leaderboard, /api/data

Pulled the surrounding code context to see the full fetch call:

curl -s '.../184-*.js' | grep -oP '.{0,200}/api/leaderboard.{0,200}'

Revealed the exact implementation:

fetch("/api/leaderboard", {
  method: "POST",
  headers: { "Content-Type": "application/json" },
  body: JSON.stringify({ timeframe: e, page: t, pageSize: n })
})

Key discovery: POST not GET. Parameters: timeframe (1/7/30), page (0-indexed), pageSize (50).

curl -s -X POST 'https://kolscan.io/api/leaderboard' \
  -H 'Content-Type: application/json' \
  -d '{"timeframe":1,"page":0,"pageSize":50}'
→ Forbidden

Forbidden — The API requires a valid browser session. Cookie/session-based protection prevents direct curl access.

Discovered the page server-renders initial data via the initLeaderboard React prop:

curl -s 'https://kolscan.io/leaderboard' \
  | grep -oP '\{[^}]*wallet_address[^}]*\}' | head -5

Extracted 616 entries from the HTML — but only the first page per timeframe, and many had missing fields due to regex limits.

By reading the full page chunk (~14KB), we found:

• Uses react-infinite-scroll-component
• Scroll target: scrollableTarget: "mainScroll" (not window!)
• Page size hardcoded at 50
• Timeframes: [1, 7, 30] → Daily / Weekly / Monthly

Installed Playwright to get a real browser session. First attempt scrolled window — captured 0 results.

// ❌ Wrong — infinite scroll listens on #mainScroll, not window
window.scrollTo(0, document.body.scrollHeight);

The infinite scroll component only triggers when #mainScroll is scrolled.

// ✅ Correct — scroll the actual container
const el = document.getElementById('mainScroll');
if (el) el.scrollTop = el.scrollHeight;

Combined with intercepting POST responses and clicking between Daily/Weekly/Monthly tabs, this captured all 1,304 entries across 472 unique wallets.

Daily:   434 entries (9 pages × 50 + 34)
Weekly:  435 entries (8 pages × 50 + 35)
Monthly: 435 entries (8 pages × 50 + 35)
─────────────────────────────────────────
Total:   1,304 entries
Unique:  472 wallets

Full data saved to JSON with GMGN-compatible import format generated automatically.