Autonomous LinkedIn PR and Lead Agent Suite

An autonomous, lightweight background daemon suite designed to run continuously on a cloud server or local machine. The suite contains two independent agents:

1. LinkedIn PR Agent (main.py): Gathers trending tech and HR news using RSS feeds, writes visionary LinkedIn post copies, searches Unsplash for matching high-resolution professional photos, and publishes them via the official LinkedIn UGC API.
2. Lead Generation Agent (lead_agent.py): Programmatically searches for HR and Talent Acquisition prospects, scores leads based on company size/role, deduplicates contacts, and emails a CSV report daily.

---

📁 Project Directory Manifest

File	Type	Description
main.py	Script	Entrypoint and scheduler for the PR Posting Agent.
lead_agent.py	Script	Entrypoint and scheduler for the Lead Generation Agent.
linkedin_api.py	Module	Handles LinkedIn REST API operations (ugcPosts, userinfo, token checks).
search_agent.py	Module	Queries news sources and RSS feeds for trending articles.
content_generator.py	Module	Calls Groq API (llama-3.3-70b-versatile) for copywriting and Unsplash image downloads.
memory_manager.py	Module	Saves posts, checks 7-day rule, adjusts weights, compiles report.
memory.json	Data	Local database containing logged posts and current topic weights.
lead_searcher.py	Module	Handles scraping operations to extract contact data.
lead_memory.py	Module	Checks for duplicate leads and stores entries in JSON/CSV formats.
lead_notifier.py	Module	Formats lead records and sends HTML summary emails.
leads.json / leads.csv	Data	Database containing extracted prospect contacts.
notifier.py	Module	Low-level SMTP client library for sending post and warning notifications.
config.py	Module	Configures credentials and environment settings.
requirements.txt	Setup	Python dependency packages requirement list.
test_linkedin_bot.py	Test	Unit tests for LinkedIn REST API calls and token expiry checks.
test_lead_agent.py	Test	Unit tests for the Lead search and log pipeline.
test_phase2.py	Test	Unit tests for topic weights tuning and SMTP report compiler.
.agents/	Config	Project-scoped guidelines and skill references.

---

🛠️ Installation & Setup

Follow these steps to set up and run the agent suite on your local machine or cloud server:

Step 1: Clone the Repository

git clone https://github.com/anu007lko/linkedin_pr_agent.git
cd linkedin_pr_agent

Step 2: Create a Virtual Environment & Install Dependencies

python3 -m venv venv
source venv/bin/activate
pip install -r requirements.txt

Step 3: Configure Environment Variables

Copy .env.example to .env and fill in your actual API tokens and SMTP credentials:

cp .env.example .env
nano .env

Ensure the following variables are configured:

• LINKEDIN_ACCESS_TOKEN: Your short-lived or long-lived LinkedIn access token.
• GROQ_API_KEY: Groq API token (uses free-tier llama-3.3-70b-versatile model).
• UNSPLASH_ACCESS_KEY: Unsplash developer access token.
• NOTIFICATION_EMAIL: Recipient address for job confirmation and warning summaries.
• SMTP_SERVER, SMTP_PORT, SMTP_USERNAME, SMTP_PASSWORD: Credentials for the sending Gmail address (requires App Passwords).

---

🧪 Run Unit Tests

Confirm everything is configured correctly by running the test suites:

# Test LinkedIn URN retrieval, token expiry logic, and UGC posting
python3 test_linkedin_bot.py

# Test dynamic weight adjustments and Unsplash downloading mocks
python3 test_phase2.py

# Test contact search APIs (Hunter, Prospeo, Snov) and lead scoring
python3 test_lead_agent.py

---

🚀 Running the Agents

Both agents run as infinite scheduler loops:

# Run the PR Posting Scheduler
python3 main.py

# Run the Lead Generation Scheduler
python3 lead_agent.py

---

☁️ Cloud VM Deployment (Systemd Services)

For continuous 24/7 execution, deploy on a Linux VM (e.g., GCP Compute Engine, AWS EC2, or DigitalOcean) using systemd daemon configurations:

1. Create PR Agent Service

Create the configuration file: sudo nano /etc/systemd/system/linkedin-pr-agent.service

Add the following configuration (replace anu007lko with your VM username):

[Unit]
Description=LinkedIn PR Posting Agent Service
After=network.target

[Service]
Type=simple
User=anu007lko
WorkingDirectory=/home/anu007lko/linkedin_pr_agent
ExecStart=/home/anu007lko/linkedin_pr_agent/venv/bin/python main.py
Restart=always
RestartSec=10

[Install]
WantedBy=multi-user.target

2. Create Lead Agent Service

Create the configuration file: sudo nano /etc/systemd/system/linkedin-lead-agent.service

Add the following:

[Unit]
Description=LinkedIn Lead Generation Agent Service
After=network.target

[Service]
Type=simple
User=anu007lko
WorkingDirectory=/home/anu007lko/linkedin_pr_agent
ExecStart=/home/anu007lko/linkedin_pr_agent/venv/bin/python lead_agent.py
Restart=always
RestartSec=10

[Install]
WantedBy=multi-user.target

3. Start and Enable Daemons

Reload systemctl configurations, enable the background processes, and start them:

sudo systemctl daemon-reload
sudo systemctl enable linkedin-pr-agent linkedin-lead-agent
sudo systemctl start linkedin-pr-agent linkedin-lead-agent

4. Monitor Daemon Status and Read Logs

# Check status
sudo systemctl status linkedin-pr-agent
sudo systemctl status linkedin-lead-agent

# Monitor real-time logs (tail -f equivalent)
journalctl -u linkedin-pr-agent -n 100 -f
journalctl -u linkedin-lead-agent -n 100 -f