agentleFS
esmailhussienesmailhussien/terraproof/skills/geospatial-frontend/SKILL.md

geospatial-frontend

Skill0 starsChanged 23 days agoMITView on GitHub

Build visually stunning geospatial demo webapps with MapLibre globe, DuckDB-WASM, and a warm dark HUD aesthetic. Use when building satellite imagery retrieval demos, embedding visualization tools, geospatial search UIs, or any map-centric data exploration frontend. Triggers on requests to scaffold a new geo demo, build a map-based UI, or create a retrieval/search visualization over satellite data.

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    No frameworks: No React, Vue, Tailwind, CSS-in-JS.

    line 13
  2. Distinctive:

    GeoJSON sources: all data as GeoJSON, expression-based paint properties for dynamic styling

    line 51
  3. Distinctive:

    Must be a LIVE demo -- simulate zoom, draw, exemplar placement, view modes

    line 69
  4. Distinctive:

    Keep files under ~500 LOC.

    line 79

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Geospatial Frontend Demo Builder

Build premium map-centric demo webapps. Derived from a production geospatial retrieval demo with battle-tested design patterns.

Architecture

  • Stack: Vite + vanilla TypeScript + MapLibre GL JS (globe) + DuckDB-WASM + pure CSS
  • No frameworks: No React, Vue, Tailwind, CSS-in-JS. Pure semantic HTML + CSS custom properties + vanilla TS.
  • Map-first: The globe IS the app. Everything else floats on top as a transparent HUD.
  • Single source of truth: Global state object; updateView() re-renders; fingerprint diffing prevents redundant DOM rebuilds.

Design System

See references/design-system.md for full token spec (colors, fonts, shadows, spacing, radii, animations, components).

Core identity: Warm espresso + ivory + rust. Dark mode only. Glassmorphed floating panels. Inspired by Anthropic Salon aesthetic.

UI Rules

See references/user-preferences.md for battle-tested patterns, confirmed features, and design rules.

Layout

viewport (100vw x 100vh, fixed)
  ::before  starfield
  ::after   vignette
  #map      MapLibre globe canvas
  brand     top-left (spinning rings + logotype)
  search    top-center (glassmorphed geocoder)
  status    top-center-below (pulsing LED + message)
  panel-L   left (query: draw controls, exemplars, results)
  panel-R   right (explore/discoveries, optional)
  tutorial  overlay + spotlight + card (z:1000+)

Panels are position: fixed, glassmorphed, with generous padding. Map fills entire viewport underneath.

Map Integration

  • MapLibre GL JS with projection: "globe" (3D globe, not flat)
  • Sentinel-2 cloudless tiles (or similar satellite basemap)
  • Layer stack: AOI fill/stroke -> draft layers -> result fill/stroke -> point circles with halos -> hover preview
  • Drawing: rectangle (shift+drag with DOM overlay) + polygon (click-to-add, double-click-to-close)
  • Fly-to animations: globe.flyTo() with curve 1.4, speed 1.2
  • GeoJSON sources: all data as GeoJSON, expression-based paint properties for dynamic styling

Data Pipeline Pattern

  • DuckDB-WASM loads Parquet manifest + shards
  • Spatial filtering via ST_Intersects or bbox containment
  • Web Worker for heavy compute (embedding distances, scoring)
  • Concurrency-limited parallel shard fetches
  • Two-phase data loading: metadata first, embeddings on demand

Tutorial System

Interactive 5-7 step spotlight walkthrough that SIMULATES real actions:

  • Box-shadow spotlight (0 0 0 9999px rgba(...)) cuts a hole over target element
  • Card auto-positioned via getBoundingClientRect(), glass-panel styled
  • Simulated clicks with animated ripple effect
  • Enter key advances; auto-place exemplars if user skips ahead
  • Store completion in localStorage
  • Must be a LIVE demo -- simulate zoom, draw, exemplar placement, view modes

Responsive

Two breakpoints only:

  • 1100px: shrink panels, hide secondary nav
  • 900px: panels collapse to bottom-center stack, max-height 40-45vh

File Organization

Keep files under ~500 LOC. Split by concern:

  • main.ts -- app state, DOM, event handlers, tutorial
  • map.ts -- MapLibre setup, layers, drawing, camera
  • types.ts -- interfaces
  • util.ts -- geometry, color interpolation, formatting
  • scoring-worker.ts -- Web Worker for compute
  • styles.css -- all CSS (pure, no preprocessor)
  • index.html -- entry with font preloads + meta tags

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.