---
title: "How an Agent Built a 3D Paris Gallery by Chaining Two Hugging Face Spaces"
description: "A coding agent built a 3D Paris gallery by chaining two Hugging Face Spaces: an image-generation Space produced dark-background monument images, and VAST-AI/TripoSplat reconstructed each single image as a 3D Gaussian splat. It then handled integration work by correcting the splats’ Y-down orientation, auto-framing monuments, compressing .ply files to .ksplat at about one-third the size, and building and deploying a static Three.js viewer. The Spaces’ agents.md files provided schemas, call and polling templates, file-upload instructions, and authentication guidance, allowing the agent to call them without a client library. The same pipeline produced six-monument galleries for Egypt and Japan after the prompts changed. The account presents callable, documented multimedia components as a way to reduce integration work and make new multimedia applications cheaper to create."
---

# How an Agent Built a 3D Paris Gallery by Chaining Two Hugging Face Spaces

[Hugging Face](https://yomu.fyi/company/hugging-face) · Mishig Davaadorj · Jun 9, 2026

**Type:** Explainer

## Summary

A coding agent built a 3D Paris gallery by chaining two Hugging Face Spaces: an image-generation Space produced dark-background monument images, and VAST-AI/TripoSplat reconstructed each single image as a 3D Gaussian splat. It then handled integration work by correcting the splats’ Y-down orientation, auto-framing monuments, compressing .ply files to .ksplat at about one-third the size, and building and deploying a static Three.js viewer. The Spaces’ agents.md files provided schemas, call and polling templates, file-upload instructions, and authentication guidance, allowing the agent to call them without a client library. The same pipeline produced six-monument galleries for Egypt and Japan after the prompts changed. The account presents callable, documented multimedia components as a way to reduce integration work and make new multimedia applications cheaper to create.

## Context

Using state-of-the-art multimedia models traditionally involved integration work such as SDKs, model weights, GPUs, input formats, polling, and other setup. The post frames documented, callable components as a building-block economy that agents can assemble, with agents.md making Gradio Spaces reachable through plain-text instructions.

## Approach / What changed

The agent chained an image-generation Space with VAST-AI/TripoSplat: prompts became monument images, which became single-image 3D Gaussian splats. It corrected Y-down output, auto-framed and compressed the files, built a Three.js viewer with scroll-to-switch and drag-to-rotate controls, deployed a static Space, and reused the pipeline for Egypt and Japan.

## Takeaways

- agents.md provides API schemas, call and polling templates, upload instructions, and authentication guidance for Gradio Spaces, so an agent can operate a Space without a client library.
- The gallery pipeline converts generated images into single-image 3D Gaussian splats, then corrects Y-down orientation, auto-frames the results, and compresses .ply files to .ksplat files that are about three times smaller.
- The same two Spaces were reused for Egypt and Japan by changing the prompts, producing six images, six splats, compression, a viewer, and a deployed Space for each country.

**Tags:** [Architecture](https://yomu.fyi/topic/architecture), [Developer Experience](https://yomu.fyi/topic/developer-experience), [Machine Learning](https://yomu.fyi/topic/machine-learning), [Open Source](https://yomu.fyi/topic/open-source)

- Source: [Hugging Face](https://huggingface.co/blog/mishig/spaces-agents-md)
- Source URL: https://huggingface.co/blog/mishig/spaces-agents-md
- Ingested by Yomu: 2026-08-29T01:50:16.038Z

[Read original post](https://huggingface.co/blog/mishig/spaces-agents-md)