SYS
LLM·gateway-down
DB
···
// composite onchain signal stack

BroScan.

WaveTrend × MoneyFlow × StochRSI fused into multi-timeframe confluence, gated by on-chain whale-flow and rug-resistance — then back-scored against real price action so you can see exactly which patterns generated edge.

24h signals
fetching
Win rate @4h
awaiting scoring
Avg return @4h
signal-favored
Whale-confirmed
fetching
// how it works

The five-stack filter.

01
Composite indicator

WaveTrend channel oscillator + smoothed MoneyFlow body × volume + StochRSI cross fused into directional bias + signal events on each candle close.

02
Multi-TF confluence

Primary signal must be confirmed by every higher TF (15m + 1h + 4h, or strict mode requiring matching signals on each). Mixed alignment is skipped.

03
Whale flow

On-chain swap data from Helius — USDC + SOL leg priced to USD. Whale net buying/selling tagged as ✓ confirms or ⚠ diverges from the signal direction.

04
Rug resistance

Mint + freeze authority (Helius) × top-10 holder concentration × Birdeye holder count / liquidity / volume floors → 0-100 score with safer / caution / risky / danger verdict.

05
Self-scoring

Every fired signal snapshots its price. A cron job back-scores 1h + 4h returns in the signal's favor and emits aggregate win-rate so you can tune what actually has edge.

// telemetry
Every alert ships with the full filter context.
See the digest
// pipe it to telegram

Set it once. Get every confluence event in your pocket.

BROBOT runs the scanner. Telegram delivers the alerts. Confluence, whale-flow, rug-score, and back-scored returns all in one message.

Get started