Skip to content

README.md

Latest commit

 

History

History
111 lines (81 loc) · 5.22 KB

README.md

File metadata and controls

111 lines (81 loc) · 5.22 KB
Raw
title Documentation
description Canonical documentation hub for get-rich-quick.
status canonical
updated 2026-04-03
tags
docs
canonical
index

Documentation

This directory now has a single job: explain the application as it exists today, separate what is implemented from what is merely planned, and keep older design research available without pretending it is the current runtime contract.

Start here

Page Use it for
Getting Started Fastest path from clone to a working local stack, first login, first strategy, and first run
Development Setup Full local development workflow, migrations, testing, frontend setup, and day-to-day commands
Architecture Audit Current architecture, safety baseline, and trading-research platform foundation status
Runbooks Operational procedures for incidents, safety controls, and routine interventions
Known Issues Current gaps, rough edges, and repo-health problems that affect operators and contributors
Roadmap Proposed next steps and the major product/engineering themes that follow from the current codebase

What the application does

get-rich-quick is a Go-based trading application with three operator surfaces:

  1. A REST API and WebSocket server for programmatic control and real-time updates.
  2. A React web UI for strategies, runs, portfolio state, memories, settings, risk, and live activity.
  3. A Cobra CLI and Bubble Tea dashboard for local control and terminal-first operations.

Underneath those surfaces is a multi-agent trading runtime that:

  1. Pulls market data, fundamentals, news, and sentiment where available.
  2. Runs analyst agents in parallel.
  3. Runs a research debate.
  4. Builds a trade plan.
  5. Runs a risk debate.
  6. Applies hard risk controls.
  7. Routes to paper or live execution adapters depending on strategy settings and broker configuration.

Feature map

Core platform

  • Strategy CRUD with schedules, paper/live mode, skip-next-run support, and typed JSON config validation.
  • Manual strategy execution through the API, UI, and CLI.
  • Persistent pipeline runs, run snapshots, agent decisions, events, conversations, audit records, orders, positions, trades, and memories.
  • JWT login plus API key authentication for protected API routes.
  • WebSocket event streaming for live run and system activity.

Trading runtime

  • Parallel analyst phase with market, fundamentals, news, and social roles.
  • Bull/bear research debate and research-manager synthesis.
  • Trader phase that generates entry, sizing, stop, and target plans.
  • Risk debate with aggressive, conservative, and neutral perspectives plus a risk-manager final signal.
  • Hard risk engine with kill switch, circuit breaker, exposure caps, and pre-trade checks.

Market coverage

  • stock strategies.
  • crypto strategies.
  • polymarket strategies as a market type in domain/runtime logic, with incomplete live execution support.

Integrations

  • LLM providers: OpenAI, Anthropic, Google, OpenRouter, xAI, Ollama.
  • Market data: Polygon, Alpha Vantage, Yahoo Finance, Binance.
  • Brokers/execution: Alpaca, Binance, local paper broker, Polymarket adapter package.
  • Notifications: Telegram, email/SMTP, Discord webhooks, PagerDuty webhooks, n8n webhooks.
  • Ops/infra: PostgreSQL, Redis, Docker Compose, Prometheus, Grafana.

Canonical vs archive

The pages below are the current source of truth for how the app actually behaves:

The rest of docs/ is still valuable, but it should be read with context:

  • ADRs record major decisions and their rationale.
  • Augr Trading Research captures strategy, execution, and risk research that informed the trading platform foundation.
  • docs/design/ contains design/spec material. Some pages still describe intended architecture rather than the exact runtime wiring that exists today.
  • Historical planning documents such as phase-*-execution-paths.md, implementation-board.md, and audit notes remain as archive material.

Reading order

If you are new to the project:

  1. Read Getting Started.
  2. Read Development Setup if you plan to contribute.
  3. Read the Architecture Audit to understand the real implementation surface.
  4. Read Known Issues before assuming every described feature is production-ready.

If you are operating the system:

  1. Read Runbooks.
  2. Review Development Setup, Architecture Audit, and the live API router in internal/api/server.go.
  3. Keep Known Issues open when debugging surprising behavior.

If you are planning future work:

  1. Read Roadmap.
  2. Review the ADRs.
  3. Use Augr Trading Research for background, not for implementation truth.