NVIDIA cuOpt — оптимизация маршрутов транспорта на Python

★ 7.6 · content

cuopt-routing-api-python is a Claude Code skill that solves vehicle routing problems — VRP, TSP, and PDP — using NVIDIA cuOpt's Python API. The skill is Python-only, as cuOpt provides no C API for routing. It covers the full problem setup via the `DataModel` class: cost and transit-time matrices, time windows, multi-dimensional fleet capacities, pickup-delivery pairs, and order precedence constraints. Built-in diagnostics through `get_status()`, `get_error_message()`, and `get_infeasible_orders()` make it straightforward to identify infeasible orders or misconfigured inputs. Reference assets include ready-to-run VRP and PDP examples. The skill is aimed at Python developers building route optimization, fleet dispatch, or last-mile delivery logic in production code.