Commently 🚀

AI-Powered LinkedIn Engagement Assistant

Commently is a smart Chrome Extension that supercharges your LinkedIn networking. It seamlessly integrates into the LinkedIn feed, allowing users to generate thoughtful, context-aware, and tone-specific comments instantly using Google Gemini.

Unlike generic comment generators, Commently utilizes a stateful memory backend, ensuring that the AI remembers your recent interactions to avoid repetitive phrasing and maintain a consistent personal brand voice.

---

📸 Demo & Screenshots

Seamless Integration	Smart Generation UI
"Commently" button injected natively into posts	Context-aware generation with tone control

n8n Backend Workflow	Result in Action
Agentic Workflow with Memory Buffer	Human-like engagement

---

✨ Key Features

• ⚡ One-Click Integration: Injects a "Generate" button directly into LinkedIn's DOM, appearing natively alongside standard interaction buttons.
• 🧠 Contextual Intelligence: Analyzes the full text of the LinkedIn post (including captions and shared articles) to generate relevant insights, not just generic praise.
• 🎨 Tone Selection: Choose from multiple personas: Professional, Funny, Supportive, Inquisitive, or Friendly.
• 📝 Custom Hints: Guide the AI with specific instructions (e.g., "Ask about their tech stack" or "Congratulate them on the new role").
• 🤖 Agentic Memory: Powered by an n8n AI Agent with a sliding window memory buffer. The AI remembers the last 3 comments you generated to ensure linguistic variety.

---

🏗️ Architecture

This project uses a hybrid architecture combining a client-side browser extension with a server-side low-code orchestration engine.

Frontend (Chrome Extension)

• Manifest V3: Modern, secure extension architecture.
• Content Scripts (content.js): Uses MutationObserver to detect new posts in the infinite scroll feed and dynamically inject the React-compatible UI. Optimized with debouncing to ensure high performance.
• DOM Manipulation: robustly handles LinkedIn's complex nested DOM structure and React-controlled input fields to extract post content and insert responses programmatically.

Backend (n8n & AI)

• Orchestration: n8n (Self-hosted/Cloud).
• LLM: Google Gemini Pro (via PaLM API).
• Logic:
  1. Webhook: Receives Post Content + User Tone + Hints.
  2. AI Agent: Processes input using a specialized System Prompt designed for human-like brevity and casual professional tone.
  3. Memory: Checks the Simple Memory node (Keyed by User ID) to recall recent outputs.
  4. Structured Output Parser: Enforces strictly formatted JSON responses to prevent hallucinations.

---

🚀 Installation & Setup

1. The Chrome Extension

1. Clone this repository.
2. Important: Copy extension/env.example.js to extension/env.js.
3. Open extension/env.js and paste your n8n Production Webhook URL.
4. Open Chrome and navigate to chrome://extensions/.
5. Toggle Developer mode (top right).
6. Click Load unpacked and select the folder extension.

2. The Backend (n8n)

1. Import the workflow file workflow/AWS n8n comment automation.json into your n8n instance.
2. Configure your Google Gemini API Credential in the "Google Gemini Chat Model" node.
3. Activate the workflow.
4. Copy the Production Webhook URL.
5. Paste this URL into the Chrome Extension's Options page.

---

💡 Usage

1. Refresh your LinkedIn feed.
2. Look for the Commently Logo (bulb icon) on any post.
3. Click the icon to open the generator.
4. (Optional) Select a Tone or add a Hint.
5. Click Generate.
6. Review the comment, click Copy or Comment to post!

---

🛠️ Tech Stack

• JavaScript (ES6+) - Frontend Logic
• Chrome Extension API - Browser Integration (Manifest V3)
• n8n - Backend Workflow Automation
• LangChain - AI Agent & Memory Management
• Google Gemini - Large Language Model
• HTML5/CSS3 - UI Components

---

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.