What problem does it solve? Running the Pywr water resources model requires operational expertise scattered across source code and practice: finding dams, converting inflow units, writing operating rules, assembling valid JSON networks, and recovering from silent failures. This Skill packages that expertise so an AI agent can execute the full reservoir operations pipeline correctly. ## Core Features & Use Cases - Eight-stage pipeline: verify installation, find dams in the GRanD database (7,320 dams), build reservoir properties, convert VIC/observed inflows, generate operating rules and demand nodes, assemble the Pywr JSON model, then run, plot, and inject regulated releases into CaMa-Flood. - Diagnostic recovery: 15 symptom-diagnosis-remedy triplets in diagnostics/triplets.yaml cover GLPK solver failures, unit conversion errors (MCM to m3, mm/day to m3/s), InfeasibleError, and silent zero-inflow runs. - HydroCraft coupling: couples with VIC (inflow), CaMa-Flood (regulated routing), DSSAT (irrigation demand), and DLBreach (dam-break trigger via overtopping checks). - Use Case: Given a basin shapefile and VIC runoff output, autonomously locate dams, build a control-curve-regulated reservoir model with environmental and irrigation demands, run a 10-year daily simulation, and validate release discharge against NSE/PBIAS convention bands. ## Quick Start Run python preflight_check.py in this directory, then ask the agent to find dams in your basin shapefile and assemble and run a Pywr reservoir model using the tools under the tools/ directory.