PostDraft

PostDraft is an agentic AI LinkedIn post generator and automated editorial reviewer. Built with LangGraph, Groq (llama-3.3-70b-versatile), a FastAPI backend, and a modern Next.js + shadcn/ui frontend.

It drafts LinkedIn posts, gathers real-time search context using Tavily when needed, evaluates drafts against a strict 7-point editorial checklist, and iterates through up to 3-5 revisions until publish-ready.

---

Key Features

• ⚡ 100% OpenAI-free: Powered entirely by LangChain Groq (ChatGroq) using ultra-fast models such as llama-3.3-70b-versatile and llama-3.1-8b-instant.
• 🔄 LangGraph Multi-Agent Architecture:
  • Writer Agent: Composes high-performing posts tailored to your selected tone, audience, and custom constraints.
  • Tavily Search Tool: Integrates real-time web search for fresh facts, statistics, and industry news.
  • Editorial Reviewer Agent: Evaluates drafts strictly against the 7 LinkedIn quality criteria:
    1. Scroll-stopping hook in first 1–2 lines
    2. One clear, valuable takeaway
    3. High skimmability (short paragraphs, 1–3 sentences each)
    4. Optimal length (~150–200 words)
    5. Engaging closing question or CTA
    6. Authentic, human voice (no robotic AI fluff)
    7. Zero hashtags
  • Automated Revision Loop: If rejected, the reviewer provides constructive feedback and the writer creates an improved draft fixing every issue.
• 🚀 FastAPI Backend:
  • Full REST API + Server-Sent Events (SSE) streaming for real-time progress updates.
  • Interactive Swagger API documentation at /docs.
• 🎨 Next.js & shadcn/ui Frontend:
  • Authentic LinkedIn desktop post preview with realistic formatting and social metrics.
  • Live agent stepper visualizing graph execution (Research → Writer → Reviewer → Revisions → Approved).
  • Reviewer rubric scorecard and attempt revision history with exact critique notes.
  • One-click copy, word count sweet spot indicator, and inline editing.
  • In-app API key and model configuration modal.

---

Tech Stack

• LLM Engine: Groq via langchain-groq (llama-3.3-70b-versatile)
• Agent Orchestration: LangGraph
• Backend: FastAPI, Uvicorn, Pydantic, SSE Starlette
• Frontend: Next.js 16 (App Router), React 19, TypeScript, Tailwind CSS, shadcn/ui design system, Lucide Icons

---

Quickstart Guide

🚀 One-Command Launch (Recommended)

Start both the FastAPI backend and Next.js frontend together with a single command:

python run.py

This starts:

• FastAPI Backend: http://127.0.0.1:8000 (Docs: http://127.0.0.1:8000/docs)
• Next.js Web UI: http://localhost:3000

---

Manual Setup & Step-by-Step Launch

1. Backend Setup

1. Create and activate a Python virtual environment:

python -m venv .venv
.\.venv\Scripts\Activate.ps1

2. Install backend dependencies:

pip install -r requirements.txt

3. Configure your environment variables in .env:

GROQ_API_KEY=gsk_your_groq_api_key_here
TAVILY_API_KEY=tvly_your_tavily_key_optional
GROQ_MODEL=llama-3.3-70b-versatile

Note: A free Groq API key can be obtained from Groq Console. Tavily search is optional; if omitted, the writer generates drafts without web search.

4. Start the FastAPI backend server:

python -m uvicorn backend.main:app --reload --port 8000

The API will be available at http://localhost:8000 with interactive docs at http://localhost:8000/docs.

---

2. Frontend Setup

1. Open a new terminal in the frontend folder:

cd frontend
npm install

2. Run the Next.js development server:

npm run dev

3. Open http://localhost:3000 in your browser.

---

3. Running via CLI (Optional)

You can also run PostDraft directly in your terminal:

python main.py --cli

Or start the FastAPI server via the CLI entrypoint:

python main.py --server

---

API Reference

GET /api/config

Returns configured services and available Groq models.

POST /api/generate

Generates a post synchronously.

Request Body:

{
  "topic": "Why simple architecture beats clever complexity",
  "tone": "Thought Leadership",
  "audience": "Tech Leaders & Engineers",
  "custom_instructions": "Focus on real-world maintenance cost",
  "use_search": true,
  "max_attempts": 3,
  "model_name": "llama-3.3-70b-versatile"
}

POST /api/generate/stream

Streams agent events via Server-Sent Events (SSE) in real time.