metadata

title: Assembler
emoji: 📚
colorFrom: blue
colorTo: blue
sdk: gradio
sdk_version: 5.18.0
app_file: app.py
pinned: false

Below is detailed documentation for the Assembler Space API, designed to be clear and precise for another LLM (or human developer) to understand and use effectively. It covers the API’s purpose, endpoints, request format, response structure, examples, and considerations. Assembler Space API Documentation Overview The Assembler Space is a Hugging Face Space that acts as a factory for creating new Hugging Face Spaces dynamically. It exposes a Flask-based API that accepts JSON input containing code, files, parameters, and configuration details. Upon receiving a valid request, it creates a new Space repository on Hugging Face, populates it with the provided files, and triggers its deployment. The API supports multiple Space types (e.g., gradio, static, docker, streamlit) and multi-file submissions, making it versatile for generating a wide range of applications. The Assembler Space itself runs as a Docker-based Space on Hugging Face, accessible via a public URL once deployed. Base URL The API is hosted at: https://-assembler.hf.space/create-space Replace with the Hugging Face username hosting the Assembler Space. The exact URL will be provided once the Space is deployed. Endpoint POST /create-space Creates a new Hugging Face Space based on the provided JSON payload. Request Format Method: POST Content-Type: application/json Body: A JSON object with the following fields: Field Type Required Description space_type String No The type of Space to create. Options: gradio, static, docker, streamlit. Defaults to gradio if omitted. files Object Yes A dictionary where keys are filenames (e.g., app.py, index.html) and values are file contents as strings. parameters Object No A dictionary of key-value pairs to be injected into Python files as PARAMS or used by the generated Space’s code. Request Constraints At least one file must be provided in files. Filenames in files should include extensions (e.g., .py, .html, .css, .txt) to ensure correct handling. space_type must match one of the supported values, or the request will fail. File contents should be valid for the intended space_type (e.g., Python code for gradio or docker, HTML for static). Response Format Content-Type: application/json Status Codes: 200 OK: Space creation succeeded. 400 Bad Request: Invalid JSON or missing/invalid fields. 500 Internal Server Error: Unexpected error during Space creation. Body: A JSON object with the following fields: Field Type Description message String A brief status message (e.g., "New Space created"). url String The URL of the newly created Space (e.g., https://huggingface.co/spaces//). note String Additional information (e.g., deployment time warning). error String (Only in error responses) Description of what went wrong. Example Requests and Responses Example 1: Static Space Request: json { "space_type": "static", "files": { "index.html": "

Hello World

Message: {{params['message']}}

", "style.css": "h1 { color: green; }" }, "parameters": { "message": "Static Space Test" } } Response (200 OK): json { "message": "New Space created", "url": "https://huggingface.co/spaces/your-username/GeneratedSpace-abc123", "note": "It may take a few minutes to build and deploy." } Notes: Static Spaces don’t automatically process parameters. The new Space’s code must handle them (e.g., via JavaScript or server-side templating if added). Example 2: Docker Space with Flask Request: json { "space_type": "docker", "files": { "app.py": "from flask import Flask\napp = Flask(name)\[email protected]('/')\ndef home():\n return f'Hello, {PARAMS['name']}'\nif name == 'main':\n app.run(host='0.0.0.0', port=7860)", "requirements.txt": "flask", "Dockerfile": "FROM python:3.10-slim\nWORKDIR /app\nCOPY . .\nRUN pip install -r requirements.txt\nEXPOSE 7860\nCMD ["python", "app.py"]" }, "parameters": { "name": "Docker User" } } Response (200 OK): json { "message": "New Space created", "url": "https://huggingface.co/spaces/your-username/GeneratedSpace-xyz789", "note": "It may take a few minutes to build and deploy." } Notes: The parameters are injected into app.py as PARAMS. The port is set to 7860 to match Hugging Face’s default. Example 3: Gradio Space Request: json { "space_type": "gradio", "files": { "app.py": "import gradio as gr\ndef greet():\n return f'Hi, {PARAMS['user']}'\ninterface = gr.Interface(fn=greet, inputs=None, outputs='text')\ninterface.launch()" }, "parameters": { "user": "Gradio Fan" } } Response (200 OK): json { "message": "New Space created", "url": "https://huggingface.co/spaces/your-username/GeneratedSpace-def456", "note": "It may take a few minutes to build and deploy." } Example 4: Invalid Request Request: json { "space_type": "invalid", "files": {} } Response (400 Bad Request): json { "error": "Invalid space_type. Must be one of ['gradio', 'static', 'docker', 'streamlit']" } Usage Example (Python) Here’s how to call the API using Python’s requests library: python import requests

url = "https://your-username-assembler.hf.space/create-space" payload = { "space_type": "static", "files": { "index.html": "

Test Page

" }, "parameters": { "key": "value" } } headers = {"Content-Type": "application/json"}

response = requests.post(url, json=payload) if response.status_code == 200: print("Success:", response.json()) else: print("Error:", response.status_code, response.json()) Additional Details Supported Space Types gradio: For interactive Python apps using the Gradio framework. Requires an app.py with Gradio code. static: For static websites. Requires at least an index.html file; supports additional files like style.css. docker: For custom apps (e.g., Flask, FastAPI). Requires a Dockerfile or uses a default one if omitted. streamlit: For Streamlit apps. Requires an app.py with Streamlit code. File Handling Python Files (.py): The parameters object is injected as a global PARAMS variable at the top of the file. Other Files: Contents are uploaded as-is; no automatic parameter injection (e.g., HTML files need custom logic to use parameters). Requirements: If requirements.txt isn’t provided, a default is generated based on space_type (e.g., gradio for Gradio, flask for Docker). Deployment Notes Build Time: New Spaces take 1-5 minutes to build and deploy on Hugging Face. The API returns immediately with a URL, but the Space won’t be live until the build completes. Port for Docker: Hugging Face expects Docker Spaces to listen on port 7860. Ensure your Dockerfile and app code align with this (see Example 2). Security Considerations The Assembler doesn’t execute the provided code; it only creates a new Space. However, ensure the generated Space’s code is safe if deploying publicly. Avoid sending sensitive data in parameters unless the new Space is private (not implemented here but can be adjusted). Limitations Rate Limits: Hugging Face may restrict frequent Space creation. Check their API documentation for quotas. File Size: Large files may fail to upload; keep contents reasonable (e.g., <1MB per file). Error Handling: Basic validation is included, but complex syntax checking isn’t performed. This documentation should equip another LLM or developer to use the Assembler Space API correctly. Let me know if you’d like to expand any section or add more examples!