🚀 LinkedIn Post Generator Agent

A modular, agent-based system for crafting engaging LinkedIn posts. This project uses Google ADK (Agent Development Kit) to orchestrate multiple specialized sub-agents, ensuring a polished, professional final post — complete with behind-the-scenes storytelling, optimized hashtags, and an optional image.

📝 Project Overview

I launched this project after noticing a clear pattern - While more people are sharing on LinkedIn, many posts end up blending together, often carrying that generic, AI-generated feel. If AI is already part of the process, why not build a tool that empowers users to create posts that stand out and authentically reflect their unique voice and goals?

This project implements a multi-phase, agent-based workflow to do just that:

✅ Phase 1: Gather post intent and additional details
✅ Phase 2: Generate a compelling behind story
✅ Phase 3: Suggest relevant hashtags
✅ Phase 4: Craft a complete post
✅ Phase 5: (Optional) Generate a prompt-based image

All phases are orchestrated by a manager agent, delegating to sub-agents for each specialized task.

📂 Project Structure

.
├── .gitignore
├── .vscode/settings.json
├── LICENSE
├── README.md
├── requirements.txt
└── linkedin_post_agent/
    ├── .env
    ├── agent.py
    ├── constants.py
    ├── prompt.py
    └── sub_agents/
        ├── story_agent/
        │   ├── agent.py
        │   └── prompt.py
        ├── hashtag_agent/
        │   ├── agent.py
        │   └── prompt.py
        ├── post_agent/
        │   ├── agent.py
        │   └── prompt.py
        └── image_agent/
            ├── agent.py
            ├── prompt.py
            └── tools/create_image.py

⚡️ Installation & Setup

Follow these steps to get the project up and running:

# Clone this repository
git clone https://github.com/sujeetgund/linkedin-post-generator-agent.git
cd linkedin-post-generator-agent

# Set up a virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install dependencies
pip install -r requirements.txt

🔑 Setting up Application Secrets

Create a .env file in the linkedin_post_agent/ directory based on the provided .env.example file:

cd linkedin_post_agent
cp .env.example .env

Google API

This project uses the Google Gemini API for generating content and images.

1. Visit the Google AI Studio.

2. Sign in with your Google account.

3. Click "Create API key" and copy it.

4. Add your Gemini API key to .env:

   GOOGLE_API_KEY=your_gemini_api_key
   

Cloudinary API

To enable image uploads, you'll need to set up Cloudinary credentials:

1. Sign up for a free Cloudinary account if you don't have one.

2. After logging in, go to your Cloudinary Dashboard.

3. Locate your Cloud name, API Key, and API Secret under the "Account Details" section.

4. Add these values to your .env file:

   CLOUDINARY_CLOUD_NAME=your_cloud_name
   CLOUDINARY_API_KEY=your_api_key
   CLOUDINARY_API_SECRET=your_api_secret
   

These credentials allow the agent to upload and manage images for your LinkedIn posts.

🚀 Running the Agent

To start the agent system, run:

python -m linkedin_post_agent

This command launches the FastAPI server, allowing you to generate LinkedIn posts through an interactive, agent-driven workflow.

For development and debugging, you can also launch the Google ADK developer UI with:

adk web

❕ Example Workflow

Here’s what you can expect from the interaction:

1. Manager agent asks: "What is the intention of your post?"
2. You provide a topic and optional details.
3. Story Agent generates an engaging backstory.
4. Hashtag Agent suggests relevant hashtags.
5. Post Agent crafts the complete LinkedIn post.
6. Optionally, Image Agent can create a relevant image for your post.
7. The final post (and image) are presented to you for review.

🤝 Contributing

Contributions, issues, and feature requests are welcome! Feel free to open an issue or submit a pull request.

📜 License

This project is licensed under the Apache 2.0 License.

🌟 Acknowledgements

• Built with the power of Google Gemini API.
• Orchestrated using Google ADK.