Run the actual mHM binary for the prepared basin. This stage must execute `KISSPATH_BINARIES/mhm/mhm` or the binary path in the run config; do not replace it with a simplified formula.
Installs into .claude/skills of the current project.
Sponsors
SSkills DirectorySkills Directory
Know which skills are safe — weekly.
SSkills DirectorySkills Directory
Is that skill safe to install?
Are you the author of Docs?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/lzwei196-docs-25be10da)
# s7_execute Skill
## Purpose
Run the actual mHM binary for the prepared basin. This stage must execute `KISSPATH_BINARIES/mhm/mhm` or the binary path in the run config; do not replace it with a simplified formula.
## Inputs
- Complete run directory from s0-s6.
- mHM binary path.
- Timeout in seconds.
- Morphology, forcing, gauge, latlon, namelist, and parameter files already present.
## Outputs
- mHM logs and model outputs under the run directory.
- `output_b1` NetCDF files such as `mHM_Fluxes_States.nc` and `mRM_Fluxes_States.nc`.
- Gauge discharge output such as `daily_discharge.out` when configured.
## Procedure
Run preflight, then execute through the stage wrapper:
```bash
python preflight_check.py
python tools/s7_execute/run_mhm.py \
--run_dir runs/wangjiaba \
--mhm_binary KISSPATH_BINARIES/mhm/mhm \
--timeout 3600
```
If the wrapper reports a model error, check `../diagnostics/triplets.yaml` before changing tools or inputs.
## Verification
Confirm that mHM finished and wrote expected output files:
```bash
find runs/wangjiaba -maxdepth 3 -type f | rg "mHM_Fluxes_States|mRM_Fluxes_States|daily_discharge|ConfigFile"
rg -n "mHM: Finished|Finished|ERROR|WARNING|L2_variable_init|POSITION_NML" runs/wangjiaba -g "*.log" -g "*.out" || true
```
If the process exits cleanly during calibration, also verify calibration artifacts in s9; Fortran `stop` can return status 0.
## Traps
- `dt_r09` in `../diagnostics/triplets.yaml`: undefined class ids or nodata leakage can produce NaN output.
- `dt_r10`: excessive L0 cell count can make runtime far slower than expected.
- `dt_v001`: near-zero gauge discharge with nonzero gridded runoff usually means a coordinate-system flag problem.
- `dt_v002`: nonzero routed field with zero gauge column can mean the gauge is on the wrong L11 cell.
- `dt_r13`: a flow-direction cycle can make startup hang after `Initialize domains ...`.
## Example
After s6 writes namelists for `runs/wangjiaba`, execute:
```bash
python tools/s7_execute/run_mhm.py --run_dir runs/wangjiaba --timeout 3600
```
