Umboni New England Weather Intelligence Platform

Umboni is a beginner-friendly weather platform for New England.

The repository has three main runtime pieces:

1. A Python fetcher that downloads free weather data from several providers.
2. A Fortran simulator that combines those provider records into a 24-hour forecast.
3. An Angular dashboard that visualizes the forecast in a browser.

The project also contains a documentation site, a self-hosted wiki, and an optional observability stack.

Start here

If you are brand new to the project, use these commands in order:

python scripts/umboni.py doctor
python scripts/umboni.py bootstrap
python scripts/umboni.py pipeline
python scripts/umboni.py run-frontend

What each command does:

• doctor checks whether the required tools are installed.
• bootstrap installs the Python and Node dependencies used by the repository.
• pipeline fetches weather data, builds the Fortran simulator, generates a forecast JSON file, and copies that forecast into the Angular app.
• run-frontend starts the Angular dashboard at http://127.0.0.1:4200.

Prerequisites

Read docs/prerequisites.md for the beginner-friendly, platform-specific setup guide.

The short version is:

• Python 3.10 or newer
• Node.js 22 or newer
• CMake 3.27 or newer
• Ninja
• GNU Fortran (gfortran)
• Doxygen
• Docker only if you want the optional wiki or observability stack

One obvious command per common task

Use the repository root for every command below.

Goal	Command
Check prerequisites	python scripts/umboni.py doctor
Install project dependencies	python scripts/umboni.py bootstrap
Fetch only the provider CSV	python scripts/umboni.py fetch
Build the Fortran simulator	python scripts/umboni.py build-fortran
Run the full fetch -> simulate -> sync pipeline	python scripts/umboni.py pipeline
Start the Angular dashboard	python scripts/umboni.py run-frontend
Run every major test suite	python scripts/umboni.py test all
Build the documentation site	python scripts/umboni.py build-docs
Serve the documentation site locally	python scripts/umboni.py serve-docs

If you already ran bootstrap, you can also use the matching npm aliases such as npm run doctor, npm run bootstrap, npm run run:pipeline, and npm run test:all.

What gets generated

After python scripts/umboni.py pipeline finishes successfully, the most important files are:

• artifacts/generated/provider-observations.csv
• artifacts/generated/new-england-forecast.json
• frontend/public/data/new-england-forecast-sample.json
• artifacts/logs/python-fetcher.log.jsonl
• artifacts/logs/fortran-simulator.log.jsonl

How to test everything

Run the full repository test flow with:

python scripts/umboni.py test all

That command runs:

• repository markdown and audit checks
• Python unit tests
• Fortran native tests
• Angular linting
• Angular unit tests
• Angular coverage tests
• Angular build
• Playwright end-to-end smoke tests
• documentation build checks

If you are on a constrained machine and want to skip the browser-based tests, use this lighter command:

python scripts/umboni.py test all --skip-end-to-end-tests

Documentation map

• Beginner home: docs/index.md
• Prerequisites: docs/prerequisites.md
• First run guide: docs/getting-started.md
• Build, run, and test guide: docs/build-run-test.md
• Troubleshooting: docs/troubleshooting.md
• Architecture overview: docs/architecture-overview.md
• Testing strategy: docs/testing-strategy.md
• External libraries: docs/external-libraries.md
• Wiki guide: see the wiki/README.md file in the repository root

Repository map

• fortran/source contains the simulation engine.
• python/new_england_weather_data_fetcher contains the provider clients, normalization logic, and the beginner task runner.
• frontend contains the Angular application.
• docs contains the GitHub Pages documentation.
• wiki contains the Wiki.js deployment and starter pages.
• observability contains Grafana, Loki, and Promtail configuration.
• samples contains deterministic fixtures and sample data.