pgm-assistantlisted
Install: claude install-skill PowerGridModel/power-grid-model-ai
# Power Grid Model Ecosystem Operations
## Package Version Status
Last reviewed: PGM v1.13.217 · PGM-DS v1.11.0 · PGM-IO v2.0.36 (checked 2026-10-30)
This skill is for operations across the full `power-grid-model` ecosystem. The repositories are tools; the user request defines the task.
Prompt starters are available in `PROMPT_LIBRARY.md`.
## Ecosystem Tool Roles
- `power-grid-model` (PGM): core calculations and validation APIs.
- `power-grid-model-ds` (PGM-DS): application-layer grid objects, graph/topology operations, mutations, and visualizer workflows.
- `power-grid-model-io` (PGM-IO): conversion between external formats and PGM-compatible datasets.
## Dependency Rules
| Constraint Category | Rule Details |
| --- | --- |
| **Reference Documents** | Start with `references/README.md`, then route to task guides in `references/tasks` and layer guides in `references/pgm`, `references/pgm-ds`, `references/pgm-io`. `REFERENCES.md` is kept as a backward-compatible landing page. |
| **Allowed Libraries** | You are strictly constrained to `power-grid-model` (PGM), `power-grid-model-ds` (PGM-DS), `power-grid-model-io` (PGM-IO), `numpy`, `pandas`, `matplotlib`, and `seaborn`. |
| **Preference** | Prefer `numpy` over `pandas` where straightforward. |
| **Graph & Visualization** | Strictly utilize any graph and network consistency logic from `power-grid-model-ds` (which exposes a `Grid` dataclass, graph algorithms, and grid visualizer tools). Do not use outside graph dependencies