quoter

Venue-agnostic market-making strategy framework — by kcolbchain (est. 2015).

The strategy core. Pluggable quoting strategies, a backtest engine, and exchange connectors. Bring your own venue.

Looking for the on-chain RWA stack (vault contracts, oracle adapter, EVM execution, risk manager)? See kcolbchain/meridian — meridian builds on quoter's strategy core and adds chain connectivity, on-chain execution, and ERC-4626 LP vaults.

What's in the box

• Strategies — constant_spread, adaptive_spread (volatility + inventory aware). Plug in your own by subclassing Strategy.
• Backtest engine — tick-driven, deterministic. CSV/Parquet export of fills.
• Connectors — base interface + reference connectors. Drop in CEX WebSockets / DEX RPC / your own venue.
• Oracle abstraction — mock + real PriceFeed source. Same interface for tests and live.
• Inventory + risk hooks — strategies see current position so they can skew quotes against directional exposure.

Quick start

git clone https://github.com/kcolbchain/quoter.git
cd quoter
pip install -r requirements.txt

# Backtest
python run.py --simulate --pair ETH/USDC --spread 0.5 --ticks 200

# Backtest with adaptive strategy
python run.py --simulate --config config/default.yaml --output fills

# Live mode (requires venue connector + keys; stub by default)
python run.py --config config/live.yaml

Strategies

Strategy	What it does
constant_spread	Fixed bid/ask spread around mid — baseline.
adaptive_spread	Spread widens with realised volatility, narrows in calm markets, skews against inventory.

Add your own by subclassing Strategy in src/strategies/. adaptive_spread.py is a clean reference shape.

Architecture

┌─────────────────────────────────────────┐
│           Strategy (pluggable)          │
│   constant / adaptive / your own        │
├─────────────────────────────────────────┤
│  Agent loop  ─►  Oracle  ─►  Connector  │
│       │              │            │     │
│       └─── Inventory + risk hooks ──────┤
├─────────────────────────────────────────┤
│             Backtest engine             │
└─────────────────────────────────────────┘

Project structure

src/
  agents/        — base_agent, rwa_market_maker (reference agent)
  strategies/    — constant_spread, adaptive_spread, your own
  oracle/        — PriceFeed abstraction (mock + live)
  connectors/    — venue interface; bring your own CEX/DEX adapter
  backtest/      — tick-driven engine + fill export
  utils/         — config, logging, inventory helpers
config/          — default.yaml, live.yaml
tests/           — pytest suite

Where this fits

• Strategy library you can drop into any venue — quoter is on its own here. The strategies and backtest engine don't assume a chain or a centralised venue.
• Reference on-chain deployment — see kcolbchain/meridian. Meridian uses quoter's strategies and ships ERC-4626 vault contracts, an oracle adapter, EVM execution, and a risk manager on top.
• Stablecoin / RWA issuance — quoter doesn't issue tokens. See kcolbchain/stablecoin-toolkit.
• Audit hardening before a strategy goes live — see kcolbchain/audit-checklist.

Running the tests

pytest -q

Contributing

See CONTRIBUTING.md and CONTRIBUTORS.md. Issues tagged good-first-issue are great entry points.

Working with kcolbchain

We build, deploy, and run market-making infrastructure for partner protocols. If you'd like to talk about a paid integration or managed market-making, see kcolbchain.com/work-with-us.

Links

• Docs: https://docs.kcolbchain.com/quoter/
• All projects: https://docs.kcolbchain.com/
• kcolbchain: https://kcolbchain.com

License

MIT — see LICENSE

---

Founded by Abhishek Krishna • GitHub: @abhicris