Build a Viral Content Prompt Factory with Make.com

Overview

This is an innovative automated AI writing workflow based on Make.com that deeply analyzes viral article writing styles, combined with large model multi-turn dialogue capabilities, to automatically generate high-quality writing prompts.

This solution achieves full automation from inspiration to finished article:

1. Style Analysis - Analyze writing styles of 5-20 viral articles
2. Prompt Generation - Generate professional writing prompts from 15 dimensions
3. Trending Capture - Auto-fetch real-time trending topics from social platforms
4. Material Integration - Search related materials via EXA API
5. Smart Judgment - Token counting judges material sufficiency
6. Article Generation - Combine prompts and materials for auto-creation

Complete workflow: Style analysis → Prompt generation → Trending capture → Material integration → Article creation

---

Core Decision Factors

When choosing automated AI writing solutions, focus on:

• Automation Level - Automation level from material collection to article output
• Prompt Quality - Whether generated prompts have depth and professionalism
• Content Originality - Whether generated content has unique perspectives
• Trending Capture Capability - Efficiency in obtaining real-time trending topics and materials

---

Technical Specifications

Specification	Value	Notes
Base Analysis Articles	5-8	Maximum up to 20
Make.com Modules	33	Current workflow core modules
Workflow Run Time	Maximum 5 minutes	Depends on material volume
Make.com Operations	100+ times	Single complete run
Prompt Definition Dimensions	15	Creative perspective, title design, article structure, etc.
Notion Databases	3	Original articles, prompts, final articles
Trending Topic Capture	30	Single capture from social platforms
EXA API Material Capture	10 data points	Internet data per capture
Material Sufficiency Threshold	10,000 Tokens	Below this won’t generate articles

---

Prerequisites

Before starting, ensure you have:

• Make.com account
• Notion account
• OpenRouter API key
• ChatGPT API key (for structured output)
• EXA API key (for material search)
• Jina API (for token counting)

Notion Database Structure

Three Notion databases: Original article library, prompt library, final article library

Need to create 3 Notion databases:

1. Original Article Library - Store viral articles to analyze
2. Prompt Library - Store generated writing prompts
3. Final Article Library - Store generated articles

Note: Original articles must be stored as code blocks, otherwise may cause data processing errors.

---

Workflow Architecture

33-module complete workflow connection diagram

Entire workflow divided into several core module groups:

1. Article Analysis Module - Fetch and analyze viral articles from Notion
2. Prompt Generation Module - Multi-turn dialogue generates 15-dimension prompts
3. Trending Capture Module - Get real-time trending topics and classify
4. Material Search Module - Search related materials via EXA API
5. Quality Judgment Module - Token counting judges material sufficiency
6. Article Generation Module - Combine prompts and materials for creation

---

Step 1: Prepare Viral Article Samples

Record 5-8 viral articles in Notion original article library:

Requirements:

• Article styles must be consistent
• Content stored as code blocks
• Can increase to 20 for higher analysis precision

Note: If article writing styles inconsistent, final generated prompts become “jack-of-all-trades”, lacking targeting.

---

Step 2: Configure Writing Style Analysis

Use OpenRouter for multi-turn dialogue, analyzing writing style from 15 dimensions:

15 Analysis Dimensions:

1. Creative perspective
2. Title design
3. Opening techniques
4. Article structure
5. Paragraph organization
6. Language style
7. Rhetorical devices
8. Emotional expression
9. Case application
10. Data citation
11. Golden sentence refinement
12. Closing techniques
13. SEO optimization
14. Interactive guidance
15. Differentiation positioning

---

Step 3: Generate High-Quality Prompts

AI-generated 15-dimension writing prompts

After multi-turn dialogue analysis, system auto-generates structured writing prompts including:

• Creative role positioning
• Title formula templates
• Structure framework suggestions
• Language style guidelines
• Rhetorical technique checklists

---

Step 4: Capture Real-Time Trending Topics

Configure social platform trending API to capture 30 real-time trending topics.

Trending Processing Flow:

1. Capture trending list
2. Use ChatGPT for classification
3. Filter suitable creation topics
4. Select 5 as writing materials

Note: ChatGPT more stable than OpenRouter for structured output; trending classification tasks recommend ChatGPT.

---

Step 5: Material Search & Quality Judgment

Search internet materials via EXA API and use Jina API to calculate token count.

Smart Judgment Logic:

• Token count >= 10,000: Materials sufficient, continue generation
• Token count < 10,000: Materials insufficient, skip this topic

This mechanism avoids generating low-quality or hollow-content articles.

---

Step 6: Auto-Generate Articles

Combine high-quality prompts and sufficient materials for AI to auto-generate viral articles.

Final generated article including title and complete structure

---

Gotchas

Special attention during setup:

1. Initial Configuration Complexity - 33 modules, 3 databases, multiple APIs; requires technical foundation

2. Time-Consuming Debugging - API Keys, JSON formats, module connection order prone to errors; requires patient debugging

3. Notion Data Sensitivity - list page content module outputs blank lines as separate “blocks”; need to limit output count

4. Trending Material Challenges - Video trending topics may lack sufficient deep materials

5. Must Save Before Running - Make.com disconnection causes unsaved workflow progress loss

6. Model Selection - OpenRouter may be inferior to ChatGPT for structured output; need to switch based on tasks

7. Material Style Consistency - Analyzed article styles must be consistent, otherwise prompts become generic

---

Use Cases

Recommended Users

• Media Teams - Need batch production of specific-style viral articles
• Content Marketing Companies - Rapidly respond to market trending topics, batch-generate marketing copy
• Tech Enthusiasts - Willing to invest time learning and configuring automation workflows
• Prompt Engineering Learners - Systematically learn prompt design by analyzing articles

May Not Suit

• Users pursuing simple “one-click generation” experience
• Users unwilling to invest time in initial configuration
• Traditional creators with extremely high originality requirements
• Budget-limited users (API calls and premium subscriptions generate fees)

---

FAQ

How many viral articles to analyze?

Recommend 5-8, maximum up to 20. Article styles must be consistent; otherwise generated prompts become too generic and lack targeting.

How long does the entire workflow take?

Based on material volume and module complexity, maximum up to 5 minutes, consuming 100+ Make operations per run.

How to judge if materials are sufficient?

Workflow uses Jina API to calculate token count. If below 10,000, materials deemed insufficient and won’t generate low-quality articles.

Is this workflow suitable for beginners?

Not very suitable. Involves 33 module configurations, 3 Notion databases, and multiple API integrations, requiring strong technical background and debugging skills.

---

Next Steps

After mastering the basic workflow, you can try:

• Adding more trending sources (Twitter, Reddit, etc.)
• Integrating multi-agent collaboration features
• Adding article quality scoring module
• Setting up auto-publishing to multiple platforms

Questions? Feel free to leave comments!