🤖 AI-Powered Content Creation Team

An autonomous multi-agent system built with the Qwen-Agent framework that researches, writes, reviews, and illustrates a complete blog post from a single topic prompt.

---

✨ Key Features

• Multi-Agent Collaboration: Utilizes a team of four specialized AI agents (Outliner, Writer, Reviewer, Image Generator) that work in sequence to produce a final article.
• Dynamic Research: The Writer Agent uses the Tavily Search API to gather real-time, relevant information, ensuring the content is timely and factual.
• Autonomous Workflow: The entire content creation pipeline is orchestrated from a single input topic, demonstrating end-to-end automation.
• Custom Tool Integration: Features a custom-built search tool (TavilySearchTool) to replace a buggy default, showcasing practical problem-solving and framework extension.
• Interactive UI: A user-friendly web interface built with Gradio allows for easy interaction without needing to use the command line.

---

🏗️ System Architecture

This project uses a pipeline architecture where the output of one agent becomes the input for the next.

[User Topic] -> [OutlineAgent] -> [Outline]
                                     |
                                     v
                  +------------------+------------------+
                  |                  |                  |
           [WriterAgent]      [WriterAgent]      [WriterAgent]  (Runs for each outline section)
                  |                  |                  |
                  +------------------+------------------+
                                     |
                                     v
                             [Combined Draft]
                                     |
                                     v
                              [ReviewerAgent]
                                     |
                                     v
                              [Polished Text]
                                     |
                                     v
[ImageAgent] -> [Image URL] -> [Final Article Assembly] -> [Save to .md file]

---

🚀 Getting Started

Follow these steps to set up and run the project locally.

Prerequisites

• Python 3.10+
• uv (or pip) for package installation
• API keys for:
  • Alibaba Cloud Dashscope (for the Qwen LLM)
  • Tavily Search API (for web research)

1. Clone the Repository

git clone [your-github-repo-url]
cd ai_content_team

2. Set Up the Environment

Create a virtual environment:

# Using uv
uv venv

# Activate it (on Windows)
.\.venv\Scripts\activate

3. Install Dependencies

uv pip install -r requirements.txt

4. Configure API Keys

Create a .env file in the project root and add your API keys:

DASHSCOPE_API_KEY="sk-your-dashscope-key"
TAVILY_API_KEY="tvly-your-tavily-key"

---

Usage

You can run the project in two ways:

1. Interactive Web UI (Recommended)

Launch the Gradio application:

python app.py

Open your browser to the local URL provided (e.g., http://127.0.0.1:7860).

2. Command-Line Interface

Run the entire workflow from your terminal:

python main.py --topic "Your chosen topic here"

The final article will be saved as a markdown file in the /outputs directory.