Photorealistic 3D Earth, public aviation and maritime signals, orbital data, earthquakes, traffic, public cameras, environmental context, and inspectable source provenance — organized as an experimental spatial-intelligence surface for XUNIA Glass Onion.
XUNIA Museum + HQ · GPT-Doug-LLM · XuniaDAO · This Repository · Upstream Project
Classification: Experimental / upstream-derived spatial-intelligence reference
XUNIA function: Glass Onion geospatial visualization and public-source correlation research
Deployment posture: Local-first browser application
Source posture: Public, attributable, inspectable data only
Operational posture: Defensive research, visualization, correlation, provenance, and human review
This repository is part of the XUNIA experimental and research ecosystem. It is not presented as a government system, classified capability, intelligence-agency service, or official product of any external vendor.
The underlying application began as the open-source God's Eye View project by Bilawal Sidhu. XUNIA preserves that upstream attribution and the repository's MIT license while using this fork as a research surface for Glass Onion architecture, public-source spatial analysis, provenance, and ecosystem integration.
Within GPT-Doug-LLM, Glass Onion is implemented as a defensive public-source correlation overlay. Its role is to combine separately sourced public evidence, create only source-grounded cross-links, verify lock state and provenance, and produce hash-sealed artifacts for review.
Glass Onion is explicitly bounded:
- no unauthorized scanning of external systems;
- no targeting or exploitation of third parties;
- no external authentication attempts;
- no automatic blocking or offensive replication;
- no modification of external systems;
- human review remains part of the evidence and publication process.
The spatial role of this repository is to provide a globe-centered interface where public signals can be visualized with their source, freshness, confidence, and modeled/live state kept distinguishable.
PUBLIC DATA SOURCES
│
▼
SOURCE ADAPTERS / NORMALIZATION
│
▼
PROVENANCE + FRESHNESS LABELS
│
▼
GOD'S EYE VIEW XUNIA
3D SPATIAL VISUALIZATION
│
▼
GLASS ONION CORRELATION LAYER
source-grounded links only
│
▼
GPT-DOUG / XUNIA ANALYSIS
│
▼
EVIDENCE / LOCK / HUMAN REVIEW
│
▼
XUNIA MUSEUM + HQ
The design goal is not to make public data look classified. The goal is to make complex open data understandable, attributable, spatially coherent, and reviewable.
The upstream application provides a browser-based 3D globe and multiple public-data layers, including:
- Live aircraft — ADS-B derived public flight telemetry and route context.
- Maritime vessels — AIS-derived ship positions when configured.
- Satellites — orbital objects propagated from public orbital elements.
- Earthquakes — public seismic events.
- Traffic — live provider data when configured, with simulated states labeled when not live.
- Public cameras — cameras published through city or transportation APIs.
- Environmental signals — fires, weather-related context, and other public datasets supported by the application.
- Infrastructure context — public geospatial infrastructure datasets where licensing permits use.
- 3D map stacks — photorealistic, aerial, and open map sources depending on configured providers.
- Sensor-style visualization — CRT, night-vision, thermal/FLIR-style, noir, and other rendering modes.
- Scene annotation — persistent marks, routes, boundaries, and measurements.
- Shareable state — camera position, layers, view modes, and tracked entities can be serialized into links.
A visual effect is not evidence. XUNIA treats the underlying source and state labels as more important than the cinematic presentation.
Every Glass Onion integration should distinguish at least these states:
| State | Meaning |
|---|---|
| LIVE | Current data retrieved from an identified public source |
| DELAYED | Real source data with a known polling or publication delay |
| RECONSTRUCTED | A derived replay or estimate built from public observations |
| MODELED | A visualization or simulation, not a live observation |
| PARTIAL | Source coverage exists but is incomplete |
| UNAVAILABLE | The source cannot currently provide data |
XUNIA does not treat modeled or reconstructed output as verified live intelligence.
Repository: https://github.com/sonoxo/gpt-doug-llm
GPT-Doug provides the broader XUNIA agent, evidence, governance, defensive-security, and Glass Onion research layers. Glass Onion's source-correlation implementation lives there and is designed around source validation, provenance-preserving links, lock verification, and bounded public-source analysis.
Repository: https://github.com/sonoxo/xuniadao
XuniaDAO acts as the machine-readable ecosystem and ontology root for XUNIA repositories. This spatial project belongs in the geospatial / mission-systems research portion of that ecosystem.
Public HQ: https://sonoxo.github.io/gpt-doug-llm/
Museum + HQ is the public index for XUNIA repositories, experiments, evidence surfaces, and project provenance. It should be treated as the public front door, while this repository remains a technical implementation and research artifact.
The current application requires Node.js compatible with the repository's package.json engine constraint:
>=24.14.0 <25 OR >=26 <27
cp .env.example .env
npm install
npm run dev -- --host localhost --port 4173Then open:
http://localhost:4173
A Google Maps API key is required for Google Photorealistic 3D Tiles. Other layers have their own provider requirements. Review .env.example, DATA_SOURCES.md, and SECURITY.md before enabling additional services.
Do not commit API keys, session tokens, secrets, or private credentials to this repository.
npm run build
npm testAdditional repository checks include:
npm run test:track
npm run qa:map-source-trayThe XUNIA fork should preserve upstream tests before introducing XUNIA-specific integration work.
Any XUNIA-specific extension added to this fork should follow these rules:
- Public-source first. Every factual layer must identify its source.
- Preserve provenance. Record source URL/provider, retrieval time, and transformation state where practical.
- Separate fact from inference. Derived conclusions must not be displayed as raw observations.
- Label simulations. Modeled, reconstructed, estimated, or synthetic layers must be visibly distinguishable from live data.
- No affiliation inflation. A public API, fork, dataset, standard, or reference repository does not establish partnership or endorsement.
- No classified claims. Do not imply access to classified, restricted, or non-public intelligence.
- No offensive automation. Glass Onion is not an exploitation, targeting, intrusion, or autonomous blocking system.
- Human review. High-impact conclusions and publication decisions remain reviewable by a person.
- Respect licenses and provider terms. Code licensing does not override third-party data, model, imagery, or API terms.
- Keep evidence inspectable. Prefer reproducible transforms, hashes, manifests, and linked source records over unverifiable assertions.
1. Acquire an identified public source
2. Validate source metadata
3. Normalize entities and coordinates
4. Attach timestamps / freshness state
5. Render the source spatially
6. Create source-grounded correlations
7. Separate observation from inference
8. Generate evidence artifacts
9. Hash / lock reviewable outputs where applicable
10. Publish only with provenance intact
This is a research and evidence workflow, not a surveillance authorization model.
This application combines code with third-party datasets, map providers, public APIs, and 3D assets. Those components do not automatically inherit the repository's MIT license.
Examples called out by the repository's existing license notice include:
- TeleGeography submarine-cable data with separate non-commercial/share-alike terms;
- OpenStreetMap-derived infrastructure data under ODbL requirements;
- NASA FIRMS material with its own public-domain/citation posture;
- runtime providers such as Google Maps, OpenSky, adsb.lol, AISStream, CelesTrak, USGS, OSM/Overpass, and municipal APIs;
- third-party 3D models under their respective individual licenses.
Before commercial deployment or redistribution, review:
Original upstream project:
https://github.com/bilawalsidhu/gods-eye-view
XUNIA fork:
https://github.com/sonoxo/gods-eye-viewXUNIA
The source code remains subject to the repository's MIT license, which identifies Bilawal Sidhu as the 2026 copyright holder. XUNIA-specific documentation and future modifications do not erase upstream authorship, copyright notices, or third-party licensing obligations.
This fork should not be interpreted as endorsement, sponsorship, partnership, or affiliation by the upstream author, data providers, government agencies, Palantir, OpenAI, NASA, NSA, or any other external organization unless independently documented by that organization.
XUNIA Glass Onion is intended to make public-source evidence easier to inspect across layers:
WHERE did this data come from?
WHEN was it observed or retrieved?
WHAT is directly supported?
WHAT is inferred or modeled?
HOW are two records connected?
CAN another reviewer reproduce the path?
That is the standard this repository should move toward as it evolves inside XUNIA.
Public sources. Provenance preserved. Inference labeled. Human review retained.