What problem does it solve? Designing a synthetic market that produces readable, non-random trading behavior from an AI agent is hard to do blind. This Skill lets you author and validate the market.json configuration that drives Jev's live paper-trading desk, so you can observe whether the agent forms coherent trading policy on your market. ## Core Features & Use Cases - Market Design: Edit market.json parameters (instrument, volatility, trend, drift, fee, style brief) and watch the OpenHarness viewer reframe Jev live without restarting. - Desk Verification: Run node "$JEV_DSH/toolchain/check.mjs" to validate the market configuration and report a verdict before calling a desk done. - Behavior Evaluation: Compare Jev's equity curve against buy-and-hold and the trend gauge to judge whether the agent buys trends and cuts losses rather than churning. - Use Case: Configure a market with strong upward drift and a momentum style brief, then observe whether Jev's BUY/HOLD/SELL actions vary with the tape instead of blindly holding. ## Quick Start Ask the agent to design a trending synthetic market in market.json, run the check toolchain to verify it, and watch Jev trade it live in the viewer.
