What problem does it solve? Tuning a tracking game so it produces a meaningful decision loop is hard: set the target too slow and the agent never misses, set it too fast and it never lands a shot. This Skill guides you through shaping a Jev Archer range (target speed, bullseye size, shot count) so the agent visibly plants slow targets and leaks fast ones. ## Core Features & Use Cases - Range Shaping: Configure speed, bullHalf, targetWidth, shots, and play style in an archer.json file to control game difficulty. - Decision Loop Design: Each tick the agent reads aim position, target position, and ticks until release, then chooses LEFT_FAST / LEFT / HOLD / RIGHT / RIGHT_FAST. - Verification Workflow: Run node "$JEV_DSH/toolchain/check.mjs" to validate the range, then observe live runs to confirm the agent plants calm settings and leaks when sped up. - Use Case: Start at speed 0.6 for a calm range, then raise speed past 1.2 or tighten the bullseye to 0.4 and report whether misses visibly increase. ## Quick Start Ask the agent to create an archer.json range with speed 0.6 and bullHalf 1.0, run the check script, then verify Jev plants the slow target and leaks when speed is raised to 1.2.
