What problem does it solve? BrainX spans multiple packages (BrainPy-State, BrainCell, BrainMass, BrainState, BrainUnit, BrainTools), and choosing the wrong scale or writing generic NumPy/JAX code leads to incorrect or non-idiomatic neuroscience simulations. This guard identifies the modeling scales a task represents and routes work to the owning package skills before any code is written. ## Core Features & Use Cases - Scale-based skill routing: Maps tasks involving point neurons, ion channels, cellular morphology, or aggregate population dynamics to the correct BrainX package skills, including multiscale combinations. - BrainX-native code enforcement: Directs implementation toward high-level APIs such as brainmass.Simulator, brainunit.math, and brainstate.transform instead of manual array machinery or raw JAX transforms on State-aware code. - Scientific validation rules: Enforces frozen parameters, independent controls, causal-branch verification, and honest claim derivation for simulation experiments. - Use Case: When asked to build a spiking network coupled to whole-brain dynamics, the guard opens both BrainPy-State and BrainMass skills, then guides a single canonical implementation using each package's orchestrator. ## Quick Start Ask the agent to build or debug a BrainX model and let this guard run first to select the right package skills and APIs.
