Alpha Leak is a production-grade, multi-phase trading intelligence system built for Pump.fun on Solana. It processes every on-chain event in real time, runs layered ML analysis across 68+ features, and executes trades with sub-millisecond inference — no polling, no delays, no separate inference server.Documentation Index
Fetch the complete documentation index at: https://mintlify.com/alphaleaks60-maker/docs2/llms.txt
Use this file to discover all available pages before exploring further.
Introduction
Understand what Alpha Leak does and how the full pipeline is organized.
Architecture
Explore the complete service map from WebSocket ingestion to live execution.
Quickstart
Connect to the live signal feed and start receiving real-time signals in minutes.
API Reference
Full reference for the SSE feed, REST endpoints, and signal payloads.
How the pipeline works
Alpha Leak is structured as four sequential phases, each building on the data produced by the one before it.Ingest
The pipeline subscribes to the Pump.fun program via WebSocket and decodes every trade, token creation, and graduation event in real time — no polling, no batching delays.
Score wallets
Phase 1 computes a 30-minute rolling feature profile for every tracked wallet — graduation rate, win rate, entry timing, return consistency — and produces a 0–100 alpha score.
Run ML inference
Phase 3 runs LightGBM models compiled to ONNX against a 68-feature vector every 5 seconds, producing calibrated probability scores for multiple price targets and time windows.
Explore by topic
The Pipeline
Ingestion, phase-by-phase processing, and how services communicate.
Intelligence
Wallet scoring, adversarial detection, genesis analysis, and market regime.
ML System
ONNX model architecture, the 68-feature vector, calibration, and training.
Live Trader
Phase-gated strategies, position sizing, circuit breakers, and monitoring.