Everything this site computes is available as plain JSON — no keys, no signup, CORS open. If you work through an AI agent (Claude, ChatGPT, Cursor, your own scripts), point it at these endpoints and it can read the scanner directly.
| Endpoint | Returns |
|---|---|
GET /api/scan | the latest full scan — one row per stock with a live setup (fields below) |
GET /api/status | data freshness: newest bar (ET), last update, current hit count |
GET /api/scans | stored scan history (newest 500) — /api/scans?id=N replays any one of them |
GET /api/chart?symbol=AAPL | full wave analysis for one symbol: pivots, count, oscillator, all system plans |
GET /api/symbols | every scanned ticker (~1,135 stocks & ETFs) |
symbol · dir (1 up / −1 down) · w4_date/w4_age (hours since the wave-4 pivot) · w5_date/w5_age — plus a plan object per system: t1/t1e (System 1: ok, pti, retrace, entry, stop, target, ps), t2/t2e (System 2: ok, q 0–4 — the trade gate is ok & q≥3), t3 (System 3: ok, long, pti), t4 (System 4: entry, stop, target, ps, cross_age). Machine-readable index file: /llms.txt.
Fetch the JSON from http://<your-site-address>/api/scan (eWaveRadar, an Elliott Wave scanner).
It returns {status, total, results:[...]}. Summarize the results as a table:
symbol, direction (dir: 1=up, -1=down), wave-4 date + age in hours, and which systems
are triggered — t1/t1e.ok = System 1 watch, t2.ok with q>=3 = System 2 trade,
t3.ok = System 3, t4 present = System 4 plan (show its entry/stop/target).
Also fetch /api/status and state the data freshness. All times are US Eastern, hourly bars.
Honest note: this runs on a home machine as a personal research project — no API key, but also no SLA. The data refreshes hourly, so polling more than once every 5 minutes gains you nothing. There is a hard limit of 120 requests/minute per IP — going over it returns HTTP 429 and a temporary block (5 minutes, escalating on repeat). Nothing here is investment advice, for you or your agent.