Die Oberfläche ist lokalisiert; Skill-Beschreibungen und -Dokumente werden per KI automatisch in Ihre Sprache übersetzt. Code, Befehle, Pfade und Fachbegriffe bleiben im Original, maßgeblich ist das Dokument des Autors.
Skills entdecken
Design und Kreativität · SKILL.md

osmnx

osmnx

Route OSMnx questions about OpenStreetMap data acquisition, graph modeling and file I/O, routing and network analysis, elevation, and static visualization.

Independent GitHub snapshot · Automated classification · Original documentation governs

Skill-Dokumentation

GitHub

name: osmnx description: "Route OSMnx questions about OpenStreetMap data acquisition, graph modeling and file I/O, routing and network analysis, elevation, and static visualization." disable-model-invocation: true metadata: disco-role: operating license: MIT

OSMnx

Use this repo skill when a task is about the osmnx package. The package helps you download, model, analyze, and visualize street networks and other OpenStreetMap features.

Read first

  • references/package-overview.md for the module map, public surface, and optional extras.
  • references/troubleshooting.md for cross-cutting install/import, optional dependency, and service-limit failures.
  • references/repo-provenance.md when you need to check whether this skill still matches the current checkout.
  • scripts/check_osmnx_environment.py --help for the bundled environment diagnostic.

Install and check

  • Install the package itself with python -m pip install osmnx.
  • If you are working from a local checkout, install it in editable mode with python -m pip install -e ..
  • To cover the optional workflows used by this skill, install the relevant extras: neighbors for nearest-node/edge search, entropy for orientation entropy, raster for raster elevation, and visualization for static plots.
  • Minimal import check: python -I -c "import osmnx as ox; print(ox.__version__)"
  • Broader environment check: python scripts/check_osmnx_environment.py

OSMnx does not expose a general end-user CLI. Use Python imports and the bundled scripts in this skill instead.

Route map

data-acquisition

Use this when the user wants to fetch OSM data or plan a query.

Covers:

  • geocode, geocode_to_gdf
  • graph_from_bbox, graph_from_point, graph_from_address, graph_from_place, graph_from_polygon, graph_from_xml
  • features_from_bbox, features_from_point, features_from_address, features_from_place, features_from_polygon, features_from_xml
  • Nominatim/Overpass settings, cache behavior, rate limits, historical snapshots, custom filters, tags, and local OSM XML fallback

Read sub-skills/data-acquisition/SKILL.md for query planning and service-limit troubleshooting.

graph-modeling-io

Use this when the user already has a graph or GeoDataFrames and needs validation, conversion, projection, simplification, truncation, or persistence.

Covers:

  • graph and GeoDataFrame validation
  • graph/GeoDataFrame conversion
  • projection helpers
  • simplification, intersection consolidation, truncation, and component selection
  • GraphML, GeoPackage, and OSM XML save/load

Read sub-skills/graph-modeling-io/SKILL.md for data-model invariants and file-format rules.

routing-analysis

Use this when the user wants nearest-node/edge matching, route solving, route GeoDataFrames, speeds, travel times, statistics, bearings, or orientation entropy.

Covers:

  • nearest_nodes, nearest_edges
  • distance helpers and edge-length preparation
  • add_edge_speeds, add_edge_travel_times
  • shortest_path, k_shortest_paths, route_to_gdf
  • street statistics, bearings, and orientation entropy

Read sub-skills/routing-analysis/SKILL.md for weighting, optional nearest-neighbor dependencies, and route-analysis troubleshooting.

elevation-visualization

Use this when the task is about node elevations, edge grades, or static plots.

Covers:

  • raster or web elevation attachment
  • edge grade calculation
  • graph, route, figure-ground, footprint, and orientation plots
  • headless plotting and image-output settings

Read sub-skills/elevation-visualization/SKILL.md for raster, matplotlib, and headless-output guidance.

Common routing clues

  • If the user mentions a city, address, boundary, building footprints, or transit stops, start with data-acquisition.
  • If the user mentions a saved graph, GeoPackage, GraphML, OSM XML, CRS, projection, simplification, or validation, use graph-modeling-io.
  • If the user mentions shortest paths, travel time, nearest nodes/edges, network statistics, or bearings, use routing-analysis.
  • If the user mentions elevation, grades, or a saved image/plot, use elevation-visualization.
  • If a task mixes workflows, complete prerequisites first and then route to the downstream sub-skill.

Runtime notes

  • Public package facts and the refresh baseline live in references/repo-provenance.md.
  • The managed router metadata lives in references/repo-routing-metadata.json.
  • The reusable smoke helpers live under scripts/ or the owning sub-skill scripts/ directories.
  • Do not link runtime instructions to the original repository checkout; use the bundled references and scripts in this skill tree.
Gespeicherte Skills