How our products are built, what ships in each release, and the decisions we make under the hood. Straight from the EasyTrading team.
Final P&L can show that an outcome changed, but rarely explains why. Trace data, decisions, orders, fills and positions to the earliest mismatch.
Read note
A profitable backtest, a successfully copied order and a live return are three different kinds of evidence. A trading workflow should never present them as interchangeable.
Read note
A smooth equity curve may come from a sound idea, parameter fitting, or fills that were never attainable. We take one backtest from hypothesis to a decision: reject it, revise it, or move it to forward validation.
Read note
Local stdio describes how an MCP client talks to the host, but it does not guarantee a fully local AI process. This guide maps the complete data path and practical security boundaries.
Read note
An AI agent may improve a strategy only on paper if the data, fees or execution model changed with the code. Here is what to preserve so another run can reproduce the result.
Read note
We compared two ways to copy a position. On one day of data they behaved identically; differences appeared after rounding and changed order-state rules.
Read note