Welcome to StarStop: an Expo-based app shell with a cosmic data pipeline that harvests planetary positions for people in a CSV and prepares those results for downstream LLM workflows.

Think of this repo as two cosmic engines working together:

• The mobile app (an Expo project) in star-stop/ — the earthly interface.
• The astro data pipeline (node scripts) under star-stop/scripts/ — the part that talks to an astrology API and stores per-person planetary data for later analysis or for sending into a model like Google Gemini.

Table of contents

• What this repo contains
• Quick start (run the app)
• Fetching celestial data (the astro pipeline)
• Output format & Gemini prep
• Security & environment
• Next steps & ideas (celestial)

What this repo contains

• star-stop/ — an Expo app scaffold created with create-expo-app. Edit app/ to change the mobile UI.
• star-stop/scripts/fetch-astro.js — a Node script that reads star-stop/data/csv/f1db-drivers-opencage-geocoded-7214233502.csv, calls a planetary-positions API for each person, and writes results to star-stop/data/astro-results/.
• star-stop/data/csv/ — CSV sources used by the pipeline.
• star-stop/data/astro-results/ — where the API responses land as per-row JSON files and an aggregate.jsonl (one JSON per line) ready for streaming into a model pipeline.

Quick start — run the Expo app

1. Install dependencies (repo root or inside star-stop/):

cd star-stop
npm install

2. Start the Expo dev server

npx expo start

Fetching celestial data (astro pipeline) This repo includes a small Node script that posts each person from the CSV to an astrology endpoint and saves the response.

1. Add your credentials to star-stop/.env.local (this file is gitignored):

export ASTRO_API_KEY='your-api-key-here'
export ASTRO_AUTH_TOKEN='your-bearer-token-here'

2. (Optional) source the file into your shell, or the script auto-loads it for you:

# from repo root — the script will auto-load star-stop/.env.local
npm --prefix star-stop run fetch-astro

# or change into the folder and run
cd star-stop
npm run fetch-astro

What the fetch script does

• Reads star-stop/data/csv/f1db-drivers-opencage-geocoded-7214233502.csv.
• For each row it sends a multipart/form POST to the astro API (/western-api/v1/planetary-positions).
• Saves each response to star-stop/data/astro-results/<safe-name>-YYYY-MM-DD.json and appends a line to star-stop/data/astro-results/aggregate.jsonl.

Output format & Gemini prep

• Per-row JSON files include the original row and the API response (status, ok, body). Example filename: Alexander_Albon-1996-03-23.json.
• aggregate.jsonl has one JSON object per line: { row, result }

This JSONL format is convenient for streaming into LLMs. If you want a ready-to-send Gemini prompt envelope, an additional helper can:

• Map/flatten response fields into a concise JSON object per person.
• Build few-shot or single-shot prompt templates that pass planetary positions and metadata as structured context to the model.

Security

• star-stop/.env.local is included in .gitignore. Do not commit secrets.
• Be cautious when pasting or sharing outputs that contain tokens or PII. Consider redacting sensitive fields before uploading or sharing.

Next steps & celestial ideas

• Add retries and exponential backoff for the astro fetcher to handle transient API errors.
• Create a scripts/prepare-gemini.js helper that converts aggregate.jsonl into a single prompt or a batch of JSON documents ready for Gemini.
• Add provenance metadata: stamp each result with request time, API version, and request id so you can audit model inputs.
• Visualize results: build an Expo screen that pulls the per-person planetary data and shows glyphs and positions in a simple sky-chart.

Enjoy the cosmos ✨