@cluesurf/view

Mapping Random Things in the World ◍

Overview

View is a set of data-curation projects that map random things going on here and there. Where the things are (data centers, mines, minerals, fabs, ports), what local systems they draw from, and the real local impact on the people nearby.

Each sub-folder of deck/ is one dataset project. The shape repeats every time: find the lists, geocode them, join to local capacity and population, then rank by local burden. The headline is never a national total. It is always a local share, because national averages hide where a single project quietly takes a town's water, power, or land.

Plain TypeScript reading and writing JSON files. No database, no app, no map frontend. Estimates ship as low / mid / high bands, and every record carries its source links and a confidence level.

Principles

• Local over national. Impact is always a share of the local system.
• Bands, not false precision. Estimates are low / mid / high.
• Confidence on everything. Source links and a confidence level per record.
• Free knowledge. Findings stay open. Data stays local where licenses require.
• Reusable pipeline. Fetch, dedup, geocode, join, rank. Same skeleton per dataset.

Layout

code/             # shared, dataset-agnostic helpers (hoisted here once a 2nd dataset needs them)
deck/             # one sub-folder per dataset project
  water-ai-complexity/
    code/         # schema, lib, fetch, build
    base/         # data: inputs + raw / clean / out (git-ignored)
    task/         # the orchestrator CLI
    note/         # the distilled snapshot for this project
    readme.md
task/             # cross-dataset CLIs
note/             # docs about the view project as a whole
package.json      # shared root for every dataset
readme.md

Quick Start

pnpm tsx deck/<dataset>/task/run.ts
# e.g.
pnpm tsx deck/water-ai-complexity/task/run.ts

Run scripts with pnpm tsx. One shared package.json, no per-dataset package.json. Source data is git-ignored under each base/ and kept local, since most sources cannot be redistributed.

Datasets

• water-ai-complexity — the water cost of US data centers. Quick read at deck/water-ai-complexity/note/00-snapshot.md. Full planning docs at note/data/water-ai-complexity/note/ in the repo root.

Starting A New Dataset

Run the plan-data-project skill (in the repo-level .claude/skills/). It walks the full workflow: research brief, planning docs, code scaffold, and the distilled snapshot.

License

MIT. See LICENSE.

ClueSurf

Made by ClueSurf, meditating on the universe ¤. Follow the work on YouTube, X, Instagram, Substack, Facebook, and LinkedIn, and browse more of our open-source work here on GitHub.