agentleFS
geoai-latamgeoai-latam/SKILL_JS5_ARCGIS/skills/arcgis-parquet/AGENTS.md

SKILL_JS5_ARCGIS / arcgis-parquet

AGENTS.md1 starsChanged 6 months agoMITView on GitHub

Quick-reference for ParquetLayer (beta). Read before loading a .parquet file. 1. GeoParquet metadata or lon/lat-style column names? Yes → new ParquetLayer({ data: new ParquetFilesData({ urls }) }). 2. WKB column without metadata? → GeometryEncodingWkb({ field }) + geometryType + spatialReference. 3. x/y columns with other names? → GeometryEncodingLocation({ xField, yField }) (x = longitude) + geometryType: "point" + spatialReference. 4. Projected CRS or ZSTD? Rewrite the file as 4326/3857 with Snappy or GZIP: the layer rejects both. 5. View? MapView…

The file

Files
AGENTS.md

Agent Guide: arcgis-parquet

Quick-reference for ParquetLayer (beta). Read before loading a .parquet file.

Quick Decision Tree

  1. GeoParquet metadata or lon/lat-style column names? Yes → new ParquetLayer({ data: new ParquetFilesData({ urls }) }).
  2. WKB column without metadata? → GeometryEncodingWkb({ field }) + geometryType + spatialReference.
  3. x/y columns with other names? → GeometryEncodingLocation({ xField, yField }) (x = longitude) + geometryType: "point" + spatialReference.
  4. Projected CRS or ZSTD? Rewrite the file as 4326/3857 with Snappy or GZIP: the layer rejects both.
  5. View? MapView / <arcgis-map> only.

Setup Checklist

  • Use data (ParquetFilesData / ParquetPortalItemData), not the deprecated urls.
  • Without GeoParquet metadata, set geometryEncoding, geometryType and spatialReference.
  • After load(): isTable is false and one queried feature has a non-null geometry.
  • The host answers HEAD and Range (206) with CORS.
  • Several files share one schema.
  • Page queries in steps of 8000 or aggregate with outStatistics.

Anti-patterns (one-liner reminders)

  • ❌ Encoding without spatialReference: every geometry null, nothing drawn, no error.
  • ❌ Trusting exceededTransferLimit: queries stop at 8000 rows and it stays false.
  • ❌ Re-exporting the file on Failed to parse file: 404, CORS, Range and HEAD all give that message.
  • ❌ where: "STATUS = 'constructed'": values are case-sensitive (0 instead of 244 149).
  • ❌ A plain { where } when the layer has a definitionExpression: start from createQuery().
  • ❌ Mixed schemas across files: Unknown field on query.
  • ❌ pandas timestamp[us]: becomes big-integer in microseconds, not a date.
  • ❌ ParquetLayer in a SceneView, or expecting applyEdits.

Files in This Skill

  • SKILL.md / SKILL.es.md — full guides
  • references/api-detail.md — ParquetLayer, data sources, encodings, field types
  • references/examples.md — GeoParquet, portal item, WKB, x/y, local file, paging, table
  • references/pitfalls.md — 12 measured pitfalls and what was not measured
  • scripts/snippet-geoparquet.ts, snippet-wkb.ts, snippet-query.ts

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.