What problem does it solve? Running geophysical inversions with pyGIMLi requires operational expertise scattered across source code, docs, and practice: converting field data formats, applying correct units, configuring regularization, and diagnosing silent failures. This Skill packages that expertise so an AI agent can reliably execute, validate, and recover pyGIMLi workflows. ## Core Features & Use Cases - Staged pipeline tools: Four executable tools cover data conversion (Res2DInv, ABEM, Syscal, SEG-2, CSV to pyGIMLi DataContainer), petrophysical parameter setup, forward/inversion runs, and output parsing to CSV/VTK. - Diagnostic recovery: 18 symptom-to-diagnosis-to-remedy triplets plus a preflight check script catch known failure modes like unit confusion (ms vs s, Ohm vs Ohm·m), duplicate electrodes, and lambda misconfiguration. - Machine-readable model identity: dag.yaml defines inputs, outputs, units, observability, and validation ranks so agents know exactly what the model predicts and how to judge results. - Use Case: Convert a CSV of ERT field measurements with electrode positions into pyGIMLi format, configure region starting models from a rock-type table, run a 2D resistivity inversion targeting chi-squared near 1, and export a coverage-masked resistivity model to VTK. ## Quick Start Run python preflight_check.py in this directory, then ask the agent to invert your ERT field data file with the run_pygimli.py tool using default regularization.
