README.md

---
title: AutoApp Builder
emoji: 🛠️
colorFrom: blue
colorTo: indigo
sdk: docker
app_port: 7860
license: mit
short_description: Generate deployable HF Spaces from a prompt.
---

# AutoApp Builder

A Docker-based Hugging Face Space that turns a product prompt into a **real,
deployable Space starter repository** — the plan *and* the source files are
generated by a Hugging Face LLM (`Qwen2.5-Coder-32B-Instruct` and fallbacks).

## What it does
- LLM designs the app: name, SDK choice, summary, features.
- LLM writes the actual files (README with Space metadata, app.py / Dockerfile /
  index.html, requirements.txt, …).
- Generated files are written safely (path-sanitized) and exported as a ZIP.
- If inference is unavailable, it falls back to a deterministic template so it
  never hard-fails.

## Configuration
- Add a secret named **`HF_TOKEN`** with the **"Make calls to Inference
  Providers"** permission (*Settings → Variables and secrets*).
- Optional override: `TEXT_MODEL`.

## Local run
```bash
docker build -t autoapp-builder .
docker run -p 7860:7860 -e HF_TOKEN=hf_your_token_here autoapp-builder
```

Then open `http://localhost:7860`.