What problem does it solve? Running the Ribasim water-resources model requires operational expertise scattered across docs and source code: correct unit conventions (m/s fluxes, m³/s flows), GeoPackage schema details, solver settings, and recovery from silent failures. This Skill packages that expertise so an AI agent can build, execute, and validate real Ribasim simulations without substituting approximations. ## Core Features & Use Cases - Network and input construction: Convert CSV/shapefile node-link definitions, basin profiles, and forcing data into a Ribasim-compatible GeoPackage database and ribasim.toml configuration with schema and unit validation. - Checked model execution: Run the actual Ribasim 2026.1.0-rc2 binary (Julia/SciML core) via a wrapper with preflight environment checks and a diagnostic triplet table mapping 18 known failure symptoms to remedies. - Output parsing and validation: Parse NetCDF results (flow_rate, level, storage, balance error, allocation) into CSVs, plots, and metrics scored against cited NSE/PBIAS convention bars. - Use Case: A hydrologist provides basin polygons, ERA5 forcing, and inflow records; the agent converts units (mm/day to m/s), builds the network, runs Ribasim, and reports flow_rate skill against the validation convention. ## Quick Start Ask the agent to run the preflight check and then build and execute a Ribasim model from your node, link, and basin CSV inputs.
