Connection setup
Everything needed to connect, test, and understand your scanner.
Deriv connection
Save changes before connecting. Reconnect after changing the ID.
Saved configuration: Current public market data. Status: disconnected.
Configured endpoint for the next connection:
Setup incomplete — save your registered App ID first.Last WebSocket endpoint actually attempted (may differ from newly saved settings):
None — Connect has not opened a WebSocket in this session.Last attempted access: No socket attempted. Current status: disconnected.
The ordinary Connect button uses public data. Use the OAuth panel below to authorize your registered application and connect an account socket. OTP credentials are redacted in diagnostics.
Getting connected
- 1. Use your registered App ID.
Copy the App ID/client_id, not the name “KarelTrade Pro”, from your Deriv dashboard. Select Current API for your OAuth application. Select Legacy v3 only if you separately registered a legacy numeric application. The old 1089 URL was an example, not your registration.
- 2. Public data does not need account authorization.
Deriv’s current public endpoint needs no App ID query parameter, OAuth login, personal access token or one-time password (OTP). Your saved API connection type selects the protocol; your ID is not used to authorize public ticks or verify the registered application. Do not paste credentials into these settings.
The OAuth panel implements Authorization Code + PKCE: redirect to Deriv with client_id, an exactly registered HTTPS redirect_uri, state and code_challenge; validate state on return and exchange the code server-side with the code_verifier. Use the resulting Bearer token for REST account discovery, then open the WebSocket URL returned by the account OTP endpoint. Never construct an account socket by appending an OAuth ID or access token to the legacy URL. OAuth Bearer requests do not require Deriv-App-ID; PAT requests do. The OAuth account connection requests the trade scope required for Options account access; the scanner still sends no trade-execution requests. Public scanning remains permission-free.
https://developers.deriv.com/docs/intro/oauth/ https://developers.deriv.com/docs/options/ws-public/ - 3. Save, connect, choose a market.
Save your ID and Current API selection, then Connect. Check the actual endpoint above, return to the scanner, select a market and wait for “LIVE TICK STREAM” and advancing tick timestamps. Switch between Standard and Pro; both use the same real feed. A connection without fresh ticks is history, not live. If contract discovery fails, the error stays visible and actionable suggestions remain blocked.
- 4. Test without execution.
Record a hypothetical condition, observe the next tick, and review your journal. In demo mode use “Next demo tick” to advance the labeled synthetic tape.
Connect your registered OAuth application
OAuth uses your saved client ID, Authorization Code + PKCE and a server-side token exchange. Account discovery requires Deriv’s trade scope, but this scanner never places orders. Authorize in a standalone tab with an exactly registered redirect URL. Tokens are held only in memory and expire; reload requires authorization again.
No authorized accounts loaded. Save the OAuth setup above and authorize to load real account choices.
Methodology & safety
Evidence before suggestions
The statistical assistant compares observed contract frequencies with neutral assumptions: uniformly distributed digits and 50% for strictly rising or falling prices. Neither assumption is a guarantee about a market.
Automatic selection compares supported contract families and digit barriers on the earliest 60% of ticks, then freezes the condition before inspecting the next 20% validation and last 20% test. All three conservative Wilson lower bounds must exceed the condition’s neutral baseline by 2.5 percentage points. The screening z increases with the number of compared markets; it is not a calibrated accuracy percentage. Sample size, verified contracts, precision, timestamp continuity and fresh-stream checks also apply. Pro requires a stricter 3.5-point excess over baseline, at least 200 outcomes in each holdout window and a more conservative screening bound, plus deeper diagnostics.
These bounds assume independent observations. Repeated scanning, autocorrelation, and selecting conditions after viewing data can invalidate statistical confidence. The displayed frequency is not a calibrated prediction of the next tick. Even a passed screen does not establish a profitable or predictive strategy.
Reliable by design
The browser connects directly to Deriv’s supported secure WebSocket market-data API. A persistent WebSocket cannot run in a short-lived server function, so it stays in the browser. Requests are limited to public market data, contract discovery, ping, and subscription cleanup.
The connection sends a heartbeat every 25 seconds and retries network interruptions with increasing delays, up to six attempts. A stream older than 10 seconds is stale and cannot produce an actionable assessment. Interrupted paper tests are unresolved, never fabricated.
To verify the wire traffic, open browser developer tools → Network → WS before clicking Connect. The current endpoint should have no app_id query. Inspect active_symbols, ticks_history and incoming tick messages; confirm symbol, quote and epoch match the screen. Disconnect and confirm the live indicator stops, then reconnect. The app does not circumvent regional restrictions. Setup and transport failures display errors without substituting demo data. Click Publish to put this draft on your live site.
Testing AUTO-SCAN
- Connect using your existing public or OAuth setup. Confirm advancing real tick timestamps.
- Open Scanner, choose Standard or Pro and select a market, or press AUTO-SCAN. Watch progress and per-market read errors. Each eligible market receives API history, contract discovery and at least two new streamed ticks.
- Inspect ranked observations and Pro final-test outcomes. No validated edge should display WAIT / NO TRADE, not a guessed digit.
- A qualified ranked market is rechecked for fresh data before automatic selection. Change the market manually at any time when market loading is idle; selection stops comparison and loads your chosen stream. Press AUTO-SCAN to resume comparisons. Switching scanner mode also stops comparison.
- Disconnect during a scan. Automatic scanning stops, connection errors are visible and retained ticks cannot qualify as live. Reconnect, then enable AUTO-SCAN again.
Comparisons run sequentially while this tab is open, with a 30-second pause between passes. They are not background server jobs or simultaneous streams for every market. Historical tests are not prospective proof; use the preserved paper journal to observe subsequent outcomes. No trades are placed.