🚀 Prompt-to-Agent Orchestrator

A zero-config, production-ready AI agent pipeline that converts a single natural language prompt into a fully executed multi-step workflow. Built with LangGraph, FastAPI, Angular, and Docker, it leverages ReAct-style planning to coordinate tools across web search, code execution, database operations, and file system tasks.

App Link: https://prompt-to-agent-orchestrator.onrender.com/

🛠️ Tech Stack

• Backend: Python 3.11, FastAPI, LangGraph, LangChain, SQLAlchemy (SQLite)
• Frontend: Angular (v19), Custom Glassmorphism CSS, RxJS
• LLM Engine: Mistral AI (mistral-large-latest)
• Infrastructure: Docker, Docker Compose, Render Blueprints (Infrastructure as Code)

---

✨ Features

• ReAct-Style Planning Agent: Uses LangGraph state machines to dynamically decide which tools to call, inspect their outputs, and iteratively arrive at the final answer.
• Integrated Multi-Tool Suite:
  • 🔍 Web Search: Custom DuckDuckGo wrapper avoiding LangChain integration issues.
  • 💻 Code Execution: Ephemeral python REPL environment.
  • 📁 File I/O: Safe workspace directory access for creating, editing, and reading files.
  • 🗄️ Database Tool: Persistent SQLite storage integration.
• Beautiful UI: Custom built Angular frontend featuring dark mode, glassmorphism design tokens, smooth animations, and structured displays for intermediate agent thought logs.
• Single-Container Deployment: Uses a multi-stage Dockerfile to build the Angular frontend and serve it statically via FastAPI.

---

📂 Project Structure

├── app/                        # FastAPI Backend
│   ├── agent/                  
│   │   ├── graph.py            # LangGraph agent state machine definitions
│   │   └── tools.py            # Custom tool implementations (Search, DB, REPL, File)
│   ├── main.py                 # FastAPI application routes (static serving & CORS)
│   └── models.py               # Pydantic data schemas
├── frontend/                   # Angular Frontend App
│   ├── src/                    
│   │   ├── app/                # UI Components & services
│   │   └── styles.css          # Global glassmorphism theme styling
│   └── package.json            
├── Dockerfile                  # Multi-stage container build definition
├── docker-compose.yml          # Local container configuration
├── render.yaml                 # Render Blueprint for zero-config cloud deploy
└── requirements.txt            # Python dependencies

---

🚀 Getting Started

Prerequisites

Ensure you have the following installed locally:

• Docker & Docker Compose
• A Mistral AI API Key (Get one at Mistral Console)

Configuration

Create a .env file in the root directory (already added to .gitignore):

MISTRAL_API_KEY=your_mistral_api_key_here
MISTRAL_MODEL=mistral-large-latest

Local Deployment (Docker)

To build and launch the application (both frontend and backend) in a single step:

# Build the container (this compiles the Angular app)
docker-compose build --no-cache

# Run the container
docker-compose up

Once running, access the application in your browser: 👉 http://localhost:8000

---

☁️ Cloud Deployment (Render)

This repository includes a render.yaml Blueprint to support instant deployment to Render.

1. Push this codebase to a personal GitHub or GitLab repository.
2. Go to Render Dashboard -> Blueprints -> New Blueprint Instance.
3. Select your repository.
4. Render will prompt you for the MISTRAL_API_KEY environment variable. Paste it in.
5. Click Apply. Render will build the Angular UI and serve it over a secure HTTPS URL automatically!

---

🔌 API Endpoints

If you wish to interact with the backend programmatically:

• POST /api/v1/execute - Send a prompt to the agent pipeline.
  • Body: {"prompt": "Your instruction here"}
  • Returns: JSON containing the final answer and a step-by-step trace of intermediate messages.
• GET /health - Health check endpoint returning {"status": "healthy"}.

Example API Request (PowerShell)

Invoke-RestMethod -Uri "http://localhost:8000/api/v1/execute" `
  -Method POST `
  -Headers @{"Content-Type"="application/json"} `
  -Body '{"prompt": "Search the web for the current price of Bitcoin."}'