Cnmaps Python Assistant

cnmaps Skill Overview

Use this guidance when helping with Python code that uses cnmaps, especially for compliant boundary data aligned with China's territorial claims, Chinese administrative boundaries, global country or region boundaries, Cartopy overlays, Matplotlib map rendering, raster masking ideas, and workflows that combine cnmaps with cartopy, matplotlib, or xarray.

When to Use This Skill

• The user mentions cnmaps directly.
• The user wants to draw China, province, city, district, or global country and region boundaries.
• The user wants to overlay administrative boundaries on a cartopy map.
• The user wants to mask, clip, or constrain raster-like scientific data to an administrative boundary.
• The user needs help deciding which cnmaps function to use for querying boundaries, plotting, clipping, or centroid labeling.
• The user needs to understand whether a returned object can be passed into cartopy, matplotlib, or a downstream masking workflow.

What Claude Should Do

• First classify the task as one of: boundary lookup, overlay plotting, raster mask/clip workflow, centroid labeling, or multi-library integration.
• Prefer real cnmaps APIs over guessed names. Use the supporting reference files before inferring missing details.
• If the user already knows several exact province, city, district, or country names, prefer one batch get_adm_maps(...) query with a list filter over many separate calls.
• When the task involves plotting, reason about the full stack: cnmaps objects, cartopy projection, Matplotlib axes, and how geometries are passed through.
• When the task involves raster clipping or masking, be explicit that cnmaps mainly provides boundary and geometry-side helpers; the full raster workflow may also involve NumPy, xarray, or Matplotlib objects.
• When the environment may not be ready, distinguish missing cnmaps, missing plotting dependencies, and missing data-provider discovery.

cnmaps Knowledge Model

Treat cnmaps as a Python function library for querying compliant boundary data aligned with China's territorial claims, retrieving geometries and centroid coordinates, drawing boundaries, and helping boundary-guided scientific plotting workflows. It is not limited to China-only maps: it also supports global country and region boundaries. Do not describe it as a CLI-first tool. Many useful tasks require combining it with cartopy, matplotlib, NumPy, or xarray rather than expecting one black-box command.

How to Work with Supporting Files

• Read references/api-overview.md when you need the package capability map.
• Read references/workflows.md when you need to choose a task-specific approach.
• Read references/return-types.md when you need to reason about what get_adm_maps or related APIs return and how those objects connect to downstream libraries.
• Read references/common-pitfalls.md when you need to avoid API hallucinations or common integration mistakes.
• Read references/capability-boundaries.md when you need to explain what cnmaps handles directly versus what still belongs to Cartopy, Matplotlib, NumPy, xarray, or external GIS tools.
• Prefer the files in examples/ when producing runnable code for users.

Output Requirements

• Prefer Python code that is directly runnable or needs only small project-specific edits.
• Briefly explain why a chosen cnmaps function or workflow fits the task.
• If the task involves cartopy, matplotlib, or xarray, include the necessary surrounding code instead of only showing the cnmaps call.
• Do not invent functions, return types, or module paths that are not confirmed.
• Do not describe cnmaps as if it were a standalone terminal command for boundary processing.

Common Failure Modes

• Treating cnmaps as a unified CLI tool instead of a Python library.
• Confusing boundary geometries with raster arrays.
• Ignoring the cartopy projection and axes context in plotting code.
• Recomputing centroids from geometry when MapRecord already exposes longitude and latitude.
• Guessing return types or function signatures without checking the supporting references.

Source: cnmetlab/cnmaps — distributed by TomeVault.