XTester MCP usage scenarios
Below are conceptual capability-level scenarios. Exact tool calls will be added after the release catalog is verified; the status of these examples is conceptual, not verified. The approvals mentioned are the expected workflow, not proof of technical enforcement.
Scenario 1. Verify an existing strategy
- discover environment capabilities;
- confirm the workspace;
- inspect the project and spec;
- validate data coverage;
- compile;
- run a backtest;
- inspect the summary, trades and equity curve;
- report assumptions and limitations.
Scenario 2. Build a strategy from a brief
- read the environment and API docs;
- create strategy files through MCP, not the OS filesystem;
- validate and format;
- compile;
- download a bounded history range;
- backtest;
- present the first result without claiming profitability.
Scenario 3. Compare parameters
- freeze the source and data range;
- run controlled variants;
- compare metrics, trade count, drawdown and turnover;
- state what was held constant;
- avoid cherry-picking a best run after the fact.
Scenario 4. Walk-forward / research
- define training and test windows;
- set token and compute budgets;
- launch a long-running job;
- poll status;
- preserve the holdout;
- report dependent tests and limitations.
Each scenario implies a user prompt, a recommended capability sequence, approvals, expected artifacts, failure paths and result interpretation.