AI coding agents have strong library defaults. Those defaults are often the packages that appear most often in training data — not the ones that keep millions of samples out of React state.
If you are building a real-time or scientific app, give the agent a public API: rekhly.chartXY({ container }).addLineSeries().setData({ x, y }). Point it at the gallery and the architecture notes. Agents write more accurate chart code when the surface is small, documented, and not a wrapper.
Related: set up Rekhly JS in five minutes.
Got any questions? I'm happy to help with products, licensing, and examples.