Vigil — The Long Watch

Vigil is an exoplanet habitability index and SETI-targeting dashboard.

It ranks confirmed exoplanets using physical habitability scores, evidence weighting, data-confidence checks, and hard guardrails for obvious false positives such as lava worlds, ultra-short-period planets, gas-giant-size bodies, and poorly measured candidates.

The project now supports two output paths:

1. A self-contained generated index.html from the Python site generator.
2. An Astro/Tailwind frontend for Netlify or any Node-capable static host.

What it does

Vigil:

• Fetches confirmed exoplanet data from the NASA Exoplanet Archive.
• Scores each planet across core habitability dimensions.
• Penalises missing data instead of letting it inflate the rank.
• Caps obviously uninhabitable worlds using orbital period, radius, and equilibrium-temperature guardrails.
• Ranks the strongest evidence-adjusted candidates by final Vigil score.
• Exports ranked candidates as JSON and CSV.
• Generates a static index.html for simple hosting.
• Builds an Astro dashboard for Netlify-style deployment.
• Falls back to built-in sample data if the full NASA scrape has not been run yet.

Scoring model

Each planet receives a 0–10 score across these dimensions:

Dimension	Weight
Magnetic field likelihood	25%
Habitable zone position	25%
Rocky surface likelihood	20%
Stellar stability	15%
System age	10%
Atmosphere retention	5%

The public ranking uses more than the raw dimension score.

Vigil Score = evidence score × data confidence factor, then physical caps

This means a planet with missing mass, orbit, atmosphere-retention, or habitable-zone data cannot float to the top just because the model skipped difficult fields.

Guardrails

Vigil applies sanity checks for:

• Ultra-short orbital periods.
• Extreme equilibrium temperatures.
• Gas-giant-size or mini-Neptune-size radii.
• Low data confidence.
• Missing habitable-zone evidence.
• Missing mass or radius data.

These guardrails were added after the first full NASA scrape exposed a failure case: a very hot, close-in planet could rank too highly if several hostile or unknown dimensions were skipped.

Project files

Vigil/
├── index.html                    # Generated standalone static website
├── scraper.py                    # Downloads confirmed planet data from NASA
├── habitability.py               # Scores planets and applies evidence/guardrail logic
├── generate_site.py              # Builds standalone index.html
├── generate_data.py              # Exports JSON and CSV data for Astro/static use
├── data/
│   ├── latest.csv                # Latest NASA scrape
│   └── ranked_candidates.csv     # Generated ranked CSV export
├── public/data/                  # Astro-readable generated JSON/CSV exports
├── src/                          # Astro frontend
├── tests/                        # Pytest scoring/guardrail tests
├── .github/workflows/            # Refresh/test/build automation
├── package.json                  # Astro/Tailwind scripts and dependencies
├── requirements.txt              # Python test dependency list
├── netlify.toml                  # Netlify build config
└── README.md

How it works

1. Scrape NASA data

python scraper.py

This downloads confirmed exoplanet data from the NASA Exoplanet Archive TAP service and saves:

data/exoplanets_YYYY-MM-DD.csv
data/latest.csv

2. Generate the standalone HTML site

python generate_site.py

This reads data/latest.csv, scores the planets using habitability.py, and writes a fresh index.html.

If data/latest.csv does not exist, the generator uses built-in sample data so the site still renders from a fresh clone.

3. Generate JSON and CSV exports

python generate_data.py

This writes:

public/data/ranked_candidates.json
public/data/top_10.json
public/data/site_meta.json
public/data/ranked_candidates.csv
data/ranked_candidates.csv

4. Run the Astro frontend locally

npm install
npm run dev

npm run dev runs python generate_data.py first so the Astro page has data available.

5. Build for Netlify

npm run build

The Netlify config uses the same command and publishes dist/.

Tests

Install the Python test dependency:

python -m pip install -r requirements.txt

Run the scoring/guardrail test suite:

python -m pytest

Current tests cover:

• Ultra-short-period score caps.
• Gas-giant-size radius caps.
• Missing-data confidence penalties.
• Temperate Earth-like candidates outranking hot close-in worlds.

GitHub Actions

The refresh workflow can be triggered manually from the Actions tab and is also scheduled to run once a year.

It runs:

python -m pytest
python scraper.py
python generate_site.py
python generate_data.py
npm install
npm run build

Then it commits refreshed generated data and index.html if they changed.

Deployment

Simple deployment options:

• Serve index.html directly with GitHub Pages or any static host.
• Use Netlify with the included netlify.toml for the Astro dashboard.
• Use any Node-capable static host that can run npm run build and publish dist/.

Why magnetic fields matter

Most habitability rankings focus heavily on distance from the star and surface temperature. Vigil gives magnetic-field likelihood equal importance to habitable-zone position because a world in the right orbit can still be sterile if stellar wind strips its atmosphere.

That makes the project less of a generic Earth-similarity list and more of a long-term survivability filter for future SETI and biosignature targets.

Data source

Planet data comes from the NASA Exoplanet Archive, using the pscomppars table from its TAP service.

NASA Exoplanet Archive: https://exoplanetarchive.ipac.caltech.edu/

Status

Functional but early.

The core scraper, scoring model, evidence-adjusted ranking logic, static HTML generator, Astro data export, and basic test coverage are now in place. The current next step is polishing the Astro UI and expanding methodology documentation.

Next improvements

• Add richer Astro filters by distance, star type, discovery method, and confidence level.
• Add a separate methodology page explaining each scoring formula.
• Add separate Habitability Score and SETI Priority Score.
• Add Earth Transit Zone and observability bonuses.
• Add downloadable CSV link in the frontend.
• Add more regression tests around known exoplanet edge cases.

License

MIT License. See LICENSE.