GitHub: formaldehid/quantforge · crates.io: quantforge
Today I'm releasing QuantForge v0.2.0.
This release moves the project from an early research/backtesting CLI into a more complete single-crate trading workflow focused on data ingestion, live bot execution, and monitoring.
The direction is still the same: Rust-first, CLI only, no UI, code-driven strategy development, small incremental open-source progress.
What is new in v0.2.0
QuantForge v0.2.0 now includes three major command groups.
data
The new data command feeds market data into a local SQLite database. Data becomes a first-class local resource that can be reused for validation, research, and live bot execution.
quantforge \
--db data/market.sqlite \
data sync \
--symbol BTCUSDT \
--interval 1mThe current sync behavior is designed to be practical for ongoing usage:
- if
startis omitted, syncing starts from now - if
endis omitted, syncing continues indefinitely
That makes data sync useful as a simple long-running feeder process.
trade
The new trade command runs the bot against live-updating data stored in SQLite. For v0.2.0, the goal was not to build a huge execution framework, but to establish the first end-to-end path:
market data → database → strategy → order execution → trade lifecycle tracking
The current implementation supports dry-run mode, live mode, strategy-driven entry and exit, market-order execution flow, position close support, and persisted bot run state.
quantforge \
--db data/market.sqlite \
trade run \
--symbol BTCUSDT \
--interval 1m \
--fast 20 \
--slow 50 \
--quote-order-qty 100 \
--mode dry-runThis is still intentionally conservative. The project is aiming for clarity and correctness first.
monitor
The new monitor command is for observing and manually managing the trading side when needed. This includes visibility into bot status, recent trades, open orders, manual order cancellation, and manual position closing.
quantforge \
--db data/market.sqlite \
monitor status \
--symbol BTCUSDTThis is important because I do not want the project to become a "black box bot." Even in a CLI-only workflow, operators need visibility and manual controls.
Current shape of the project
QuantForge is still not trying to be a polished end-user platform. It is becoming a code-first foundation for market data ingestion, backtesting, strategy experimentation, controlled live execution, and operational monitoring.
That means the project remains best suited for developers who want a Rust-native CLI framework they can inspect, modify, and extend directly.
What is next
For the next iterations, I want to keep improving the reliability and ergonomics of the workflow:
- stronger reconciliation between local run state and exchange state
- better safety rails around live execution
- more built-in strategies and indicators
- improved testing around exchange behavior and order lifecycle
- clearer contributor workflows and release discipline
If you want to follow the project or try it locally, take a look at the repository and feel free to open issues or ideas.
v0.2.0 — live trading, data sync & monitoring