Claude Skills by lzwei196

Use the real `tools/summarize_rainfall.py` workflow. Never replace it with a mental calculation or claim an output that was not written. Validate inputs, run the tool, inspect the output CSV, and report failures with their exact fix. This KI turns a daily precipitation CSV into monthly totals. It is a task workflow, not a physical process simulator.

**Package**: `hydrocraft-adcirc-ocean` v1.0.0 **Model**: ADCIRC v56.2.1 — 2D/3D finite element circulation model **Created by**: Jianyun Zhang Research Group, Hohai University **Last updated**: 2026-03-26 **Stats**: 4 tools | 7 skill documents | 20 diagnostic triplets | ~1,600 lines of validated Python **Validation status**: `build_tested` ---

> **MANDATORY EXECUTION POLICY** — READ BEFORE PROCEEDING > > You MUST run the **actual model binary or package** described in this document. > If the model fails to import, compile, or execute, you MUST: > 1. Check `diagnostics/triplets.yaml` for a matching error pattern > 2. Try the fix suggested in the triplet's `remedy` section > 3. If still failing, report the error to the user with full details > > You MUST NOT substitute a simplified Python formula, regression equation, > or hand-coded...

**Model**: APEX v0806 (PE32 Windows binary via Wine) **Distributor**: Texas A&M AgriLife / Blackland Research and Extension Center **Source**: https://epicapex.tamu.edu/software/ **Language**: Fortran 90 (Intel Fortran compiled, PE32 executable) **Domain**: Field- and watershed-scale agronomy / hydrology / water-quality **Validation status**: `validated` — Bengbu China corn 6.21 t/ha vs 5.6 observed (+11% bias). Multi-subarea farm (4 fields) validated with CMFD weather, HWSD soil, GGCMI calen...

> **MANDATORY EXECUTION POLICY** — READ BEFORE PROCEEDING >

- **Package**: hydrocraft-alpine3d-snow - **Version**: 1.0.0 - **Model**: Alpine3D (with MeteoIO + SNOWPACK) - **Domain**: Cryosphere — spatially distributed snow, energy balance, and runoff in mountainous terrain - **Created**: 2026-03-26 - **Last updated**: 2026-04-30 - **Tools**: 4 | **Skill Documents**: 6 | **Diagnostic Triplets**: 22 | **Validation**: T3 (3 SNOTEL sites, 5 WY each)

**Package**: `hydrocraft-amanzi-ats` v1.0.0 **Model**: Amanzi + ATS (Advanced Terrestrial Simulator) **Domain**: Groundwater flow, reactive transport, integrated surface-subsurface hydrology **Created by**: Auto-dissection pipeline **Last updated**: 2026-03-25 **Stats**: 4 tools | 5 skill documents | 18 diagnostic triplets | ~1,200 lines of validated Python **Validation status**: `NOT_RUNNABLE` (binary requires MPI+Trilinos build; KI tools validated with Python surrogate only) ---

**Package**: `aquacrop-ospy-knowledge` v1.0.0 **Target Model**: AquaCrop-OSPy v3.0.12 (FAO AquaCrop v7.1 Python implementation) **Domain**: Crop-water productivity modeling and deficit irrigation optimization **Install**: `pip install aquacrop` (requires numpy>=1.22, pandas>=2.0, tqdm>=4.65) ---

> **Stage ID**: s10_water_productivity > **Pipeline order**: 10 of 10 > **Depends on**: s9_output_analysis

> **Stage ID**: s1_crop_selection > **Pipeline order**: 1 of 10 > **Depends on**: none

> **Stage ID**: s2_soil_profile > **Pipeline order**: 2 of 10 > **Depends on**: none

> **Stage ID**: s3_weather_prep > **Pipeline order**: 3 of 10 > **Depends on**: none

> **Stage ID**: s4_initial_conditions > **Pipeline order**: 4 of 10 > **Depends on**: s2_soil_profile

> **Stage ID**: s5_irrigation > **Pipeline order**: 5 of 10 > **Depends on**: none

> **Stage ID**: s6_field_management > **Pipeline order**: 6 of 10 > **Depends on**: none

> **Stage ID**: s7_model_assembly > **Pipeline order**: 7 of 10 > **Depends on**: s1_crop_selection, s2_soil_profile, s3_weather_prep, s4_initial_conditions, s5_irrigation, s6_field_management

> **Stage ID**: s8_execution > **Pipeline order**: 8 of 10 > **Depends on**: s7_model_assembly

> **Stage ID**: s9_output_analysis > **Pipeline order**: 9 of 10 > **Depends on**: s8_execution

> **Model**: BIOME-BGC 4.2 (University of Montana NTSG) > **Domain**: Terrestrial biogeochemistry -- daily C/N/water cycling for forests, grasslands, shrublands > **Pipeline stages**: 8 | **Tools**: 6 | **Skill documents**: 3 | **Diagnostic triplets**: 25 > **Binary**: `KISSPATH_BINARIES/biome-bgc/bgc-src/bgc` > **Status**: binary_only (bundled example data validated 2026-03-24)

Generate the BIOME-BGC initialization (.ini) file that controls all aspects of a simulation: input file paths, simulation timing, climate change scenarios, CO2, site physical properties, initial conditions, and output configuration.

Convert CMFD, MSWX, or VIC forcing data to BIOME-BGC daily meteorological input format. This stage contains the most dangerous unit conversion traps in the entire pipeline.

Run BIOME-BGC for hundreds to thousands of simulated years to bring soil carbon and nitrogen pools to steady-state equilibrium. This is the most critical and unique stage in the BIOME-BGC pipeline -- no other HydroCraft model requires multi-century spinup.

**Package**: `hydrocraft-bmi-framework` v1.0.0 **Model**: BMI v2.0 — Basic Model Interface Specification **Source**: https://github.com/csdms/bmi **Created by**: CSDMS (Community Surface Dynamics Modeling System), University of Colorado Boulder **Authors**: Eric W.H. Hutton, Mark D. Piper, Gregory E. Tucker **Last updated**: 2026-03-26 **Stats**: 4 tools | 5 skill documents | 17 diagnostic triplets | ~1,200 lines of validated Python **Validation status**: `specification_validated` (bmi-exampl...

**Package**: `hydrocraft-caesar-lisflood` v1.0.0 **Model**: HAIL-CAESAR v1.0 (High-performance Architecture Independent LISFLOOD-CAESAR) **Domain**: Geomorphology / Flood inundation / Landscape evolution **Language**: C++ (compiled with g++, OpenMP parallelisation) **Source**: https://github.com/dvalters/HAIL-CAESAR **Created by**: Hydrocraft dissection pipeline **Last updated**: 2026-03-26 **Stats**: 4 tools | 5 skill documents | 15+ diagnostic triplets **Validation status**: `test_validated...

**Package**: `hydrocraft-cequalw2-reservoir` v1.0.0 **Model**: CE-QUAL-W2 v4.5 (2D laterally-averaged hydrodynamic and water quality) **Created by**: Jianyun Zhang Research Group, Hohai University **Last updated**: 2026-03-24 **Stats**: 16 tools | 6 skill documents | 25 diagnostic triplets | 3,434 lines of validated Python **Validation status**: `binary_only` (pending source compilation and example validation) ---

Collect user inputs and make key decisions about the CE-QUAL-W2 simulation setup. This includes reservoir selection, simulation period, forcing source, grid resolution, and whether to enable water quality constituents.

Run the CE-QUAL-W2 binary with comprehensive preflight checks and post-run validation. CE-QUAL-W2 reads `w2_con.npt` from the current working directory and is single-threaded.

Parse CE-QUAL-W2 output files and generate visualizations. The signature output is the **curtain plot**: a 2D longitudinal-vertical cross-section showing temperature (or other variables) from upstream to dam at different depths.

Convert reservoir geometry into a CE-QUAL-W2 segment-layer grid and write the bathymetry file (bth_wb*.npt). This is the most complex and labor-intensive stage. For HydroCraft automation, two modes are supported: DEM-based (for any reservoir with DEM coverage) and idealized (for quick starts with minimal data).

Convert HydroCraft forcing data (CMFD/MSWX/VIC) to CE-QUAL-W2 met file format. This stage performs the most dangerous unit conversions in the entire pipeline. Getting these wrong produces NO error message — just silently wrong water temperatures.

Assemble the CE-QUAL-W2 master control file `w2_con.npt` from all upstream tool outputs. This is the most error-prone stage because w2_con.npt uses **8-character fixed-width fields** read by Fortran column position. A single misalignment corrupts all downstream values silently.

**Package**: `hydrocraft-cism-icesheet` v1.0.0 **Model**: CISM v2.1 (Community Ice Sheet Model) **Domain**: Cryosphere -- land ice dynamics (ice sheets, ice shelves, glaciers) **Language**: Fortran 90 with CMake build system **Created by**: Knowledge Dissection Toolkit **Last updated**: 2026-03-26 **Stats**: 5 tools | 7 skill documents | 20 diagnostic triplets | ~1,500 lines of validated Python **Validation status**: `synthetic_validated` (dome test case) ---

**Package**: `hydrocraft-classic-biogeochem` v1.0.0 **Model**: CLASSIC (CLASS v3.6.2 + CTEM v2.0) **Created by**: Knowledge Dissection Toolkit **Last updated**: 2026-03-26 **Stats**: 4 tools | 5 skill documents | 18 diagnostic triplets | ~2,000 lines of validated Python **Validation status**: `dissected` ---

Convert global reanalysis or site-level meteorological observations into CLASSIC-compatible netCDF forcing files. CLASSIC requires **7 separate netCDF files** (one variable per file) with a specific time encoding that differs from standard CF conventions.

Build a CLASSIC initialization netCDF file containing soil texture, vegetation parameters, and initial values of all prognostic variables. This is the most complex input file, with ~50+ variables spanning soil, vegetation, carbon, and model state.

Configure the CLASSIC job options file and select the appropriate run parameters namelist. The job options file is the master control file that determines what the model simulates, which input files it reads, and what output it produces.

Compile and run the CLASSIC model binary. CLASSIC is a Fortran model compiled with gfortran (serial) or mpif90 (parallel), linked against the netCDF-Fortran library.

Extract CLASSIC netCDF output variables into CSV format for analysis, plotting, and comparison with observations. Compute summary diagnostics including carbon budget balance, energy closure, and water balance.

- **Package**: CLM5___CTSM Knowledge Infrastructure v1.0 - **Model**: Community Land Model 5 / Community Terrestrial Systems Model (CTSM 5.4) - **Domain**: Biogeochemistry, Land Surface, Hydrology, Carbon-Nitrogen Cycling - **Created**: 2026-03-26 - **Tools**: 4 Python scripts - **Validation**: Documented ---

Configure a CLM5/CTSM simulation by selecting the appropriate compset (component set), resolution, machine, and physics options. This stage determines the entire model behavior — incorrect configuration leads to wrong science or build failures downstream.

Convert atmospheric reanalysis or observational data into the format required by CLM5's data atmosphere component (DATM/CDEPS). This is the most error-prone stage due to unit conversion requirements — incorrect forcing units produce silent errors that propagate through all model outputs.

Map external soil property databases (HWSD, SoilGrids, or custom surveys) to CLM5's 25-layer soil column structure. CLM5 uses sand/clay texture percentages, organic matter density, and color class to determine hydraulic conductivity, thermal properties, and water retention via pedotransfer functions.

Build and execute the CLM5/CTSM model binary, either through the CIME case management system or standalone via LILAC. This stage handles compilation, job submission, runtime monitoring, and error capture.

Extract, analyse, and validate CLM5 history file outputs. Convert NetCDF history files to analysis-friendly formats (CSV, JSON), compute summary statistics, and compare against observations or published reference values.

| Field | Value | |-------------------|--------------------------------------------------------------| | **Package** | COAWST KI v1.0 | | **Model** | COAWST v3.8 (Coupled-Ocean-Atmosphere-Wave-Sediment Transport) | | **Authors** | John C. Warner (USGS), Brandy Armstrong, Ruoying He, Jesse Maitland | | **Repository** | https://github.com/DOI-USGS/COAWST ...

**Package**: `cosipy-ki` v1.0.0 **Model**: COSIPY v2.0.2 **Source**: https://github.com/cryotools/cosipy **Domain**: Cryosphere — glacier and snowpack energy/mass balance **Last updated**: 2026-03-26 **Stats**: 4 tools | 6 skill documents | 20 diagnostic triplets | ~2,400 lines of validated Python **Validation status**: `validated` (Zhadang Glacier, Tibet, ERA5 2009) ---

**Package**: `hydrocraft-crest-ef5` v1.1.0 **Model**: EF5 v1.2.3 with CREST water balance + Linear Reservoir / Kinematic Wave routing **Framework**: Ensemble Framework For Flash Flood Forecasting (EF5) **KDT version**: 5.1.2 (uses `ki_tools_common` for forcing/metrics/cross-platform) **Created by**: Jianyun Zhang Research Group, Hohai University **Last updated**: 2026-04-28 (added Stage-1 `prepare_basic_grids` tool; corrected `ef5 -p` documentation) **Stats**: 5 tools | 5 skill documents | 18...

**Package**: hydrocraft-crhm v1.0.0 **Model**: CRHM 1.3 (crhmcode, University of Saskatchewan) **Domain**: Cold regions hydrology -- blowing snow, sublimation, frozen soil infiltration, energy-balance snowmelt **Coupling target**: VIC 5.1.0 (cold regions process enhancement) ---

> **Stage ID**: s1_basin_setup > **Pipeline order**: 1 of 6 > **Depends on**: none

> **Stage ID**: s2_observation_data > **Pipeline order**: 2 of 6 > **Depends on**: none (can run in parallel with s1)