SimpleSimulationEngine

Before writing new code — search existing implementations, topical audits, READMEs, file-header caveats

After implementing — update file headers, README, topical audits; note duplication

Navigate and debug OrbitalWar / SpaceCraft — design, truss sim, combat, radiosity, orbits; hub to docs and tests

Diagnostic plots/visualizations — shared utilities, output paths, plot style conventions

Dedicated documentation work — OKF format, topical audits, extract-overview and inline-doc workflows

Writing new code — inventory-first, module-vs-script placement, no-new-files, STOP triggers for duplication

Building C++ (FireCore/SPFF) — build dirs, cmake targets, ASAN vs opt modes

Fixing C++ memory crashes (double-free, segfault, UAF) and ASAN

Optimizing CPU compiled code (C/C++/Fortran/Rust) — cache hierarchy, SIMD, data-oriented design, loop optimization, memory layout

Python↔C/C++ ctypes bindings — auto-build, zero-copy buffers, memory layout

Implementing/debugging forcefields (UFF, SPFF) — invariants, switch semantics, buffer parity

Debugging GPU/OpenCL/CUDA kernels — gated prints, CPU↔GPU tracing, barrier pitfalls

Optimizing GPU/OpenCL kernel performance — memory hierarchy, gather vs scatter, tiling, auxiliary arrays, local memory reductions, kernel launch overhead

Comparing two calculations — reference vs conservation-law parity, bisect strategy, float32/64 tolerance

Porting Fortran/C++→PyOpenCL — OpenCLBase, kernel caching, persistent buffers, GPU performance

Performant Python for scientific computing — vectorization, NumPy anti-patterns, preallocation

Running/writing tests — pytest classes, markers, ref_data regression, debug output paths