market

simple & extensible prediction markets built with rust

features

• lmsr pricing - polymarket-style logarithmic market scoring rule with unlimited liquidity
• instant trading - buy and sell shares at algorithmically determined fair prices
• oracle resolution - designated resolvers ensure accurate market outcomes
• price charts - real-time probability tracking with historical data
• multi-theme ui - choose from light, dark, hacker, sepia, and pastel themes
• session management - secure cookie-based authentication

tech stack

• rust - blazing fast & memory safe
• axum - modern web framework
• sqlite - embedded database via sqlx
• askama - type-safe html templates
• chart.js - interactive price charts
• htmx - dynamic ui without heavy javascript

quick start

using docker (recommended)

# pull and run the latest image
docker run -d \
  -p 3000:3000 \
  -v market-data:/app/data \
  ghcr.io/ludeed/market:latest

# open browser
open http://localhost:3000

building from source

# clone the repo
git clone https://github.com/ludeed/market.git
cd market

# build and run with docker
docker build -t market .
docker run -p 3000:3000 -v market-data:/app/data market

# or run directly with cargo
cargo run

how it works

creating markets

anyone can create a binary prediction market with:

• a yes/no question
• an end date
• an optional oracle (resolver)

markets use lmsr (logarithmic market scoring rule) for pricing, which provides:

• unlimited liquidity (no liquidity pools to drain)
• fair pricing based on outstanding shares
• smooth price discovery

trading

buy or sell shares at any time:

• see real-time cost preview before trading
• track your positions and p&l
• view historical price charts

resolution

when a market ends:

• the designated oracle (or creator) resolves the outcome
• winning shares pay out $1 each
• losing shares are worthless
• profits are automatically credited

project structure

market/
├── src/
│   ├── domain/          # core business logic (lmsr, positions, markets)
│   ├── repository/      # database access layer
│   ├── web/            # http handlers, routing, sessions
│   └── db/             # database utilities
├── templates/          # askama html templates
├── static/            # css and client-side assets
└── migrations/        # sql schema

configuration

the app uses environment variables:

DATABASE_URL=sqlite:market.db  # database path
HOST=0.0.0.0                   # bind address
PORT=3000                      # port number

development

# run with auto-reload
cargo watch -x run

# run tests
cargo test

# format code
cargo fmt

# lint
cargo clippy

roadmap

see ROADMAP.md for planned features and known issues

credits

built by ludeed & claudius

license

MIT