AI Blog Content Machine
Research-backed, fully automated blog publishing. From topic selection to a live, SEO-optimised post with editorial images — in under 4 minutes.
Free download · enter your email · n8n workflow JSON
How it works
Six steps, zero manual work
The entire pipeline runs inside n8n. Each step is a discrete node — observable, retryable, and replaceable without touching the rest of the workflow.
Topic Selection
Every 3 minutes, the agent selects 5 topics from a curated pool of 25 across five category tracks: AI Agents & Frameworks, AI Industry Impact, AI Tools & Productivity, Latest AI Models, and AI Development Progress.
25 topics · 5 categories · runs every 3 minLive Research
Each topic is sent to Serper.dev for dual-channel retrieval — 8 top Google organic results for depth, plus 6 live news results for freshness. Citation sources are extracted and structured before writing begins.
Serper Web API + News API · 14 sources per postSEO Strategy
Claude Sonnet 4.6 analyses the competitor results and constructs a full E-E-A-T SEO strategy: primary keyword, 8 secondary keywords, target audience, unique angle, 60-char title, 155-char meta, H1, and a section-by-section outline with word targets.
Claude Sonnet 4.6 · competitor gap analysisContent Generation
Claude Opus 4.8 writes the full article — minimum 4,000 words, structured HTML, 8+ in-text citations, comparison tables, a 5-question FAQ section with 100–150 word answers, and a full references list. Every statistic is sourced.
Claude Opus 4.8 · 16,000 token context · full HTML outputImage Generation
Claude Sonnet generates tailored image prompts for the article. FAL.ai Flux Pro renders the cinematic 16:9 featured image; Flux Dev renders two contextual in-article images — all without text or logos, editorial quality.
FAL.ai Flux Pro (featured) + Flux Dev (2× in-article)Auto-Publish
Claude assigns the post to the best category (or creates a new one). The full post — HTML, metadata, images, citations, schema markup, tags, and read time — is saved to Supabase and goes live immediately with no human intervention.
Supabase · category AI · live on publishRequirements
What you need to run it
Four external credentials are required. All four offer free tiers sufficient for testing. Estimated cost at full production speed (~100 posts/day): $15–25 USD/day.
Anthropic API Key
ANTHROPIC_API_KEYPowers two Claude models: Sonnet 4.6 for SEO strategy and image prompts (fast, cost-efficient); Opus 4.8 for full article generation (maximum quality, 16k tokens). One key covers both.
Serper API Key
SERPER_API_KEYProvides real-time Google Search and Google News results via the Serper.dev API. Used for competitor research and live news retrieval before every article is written.
FAL.ai API Key
FAL_KEYGenerates all images via the FAL.ai platform. Flux Pro renders the cinematic featured image; Flux Dev renders two editorial section images. All images are 16:9 landscape JPEG.
Supabase Project
SUPABASE_URL + SUPABASE_SERVICE_ROLE_KEYStores the complete post record: HTML content, SEO fields, image URLs, citations, tags, category ID, schema markup, read time, and word count. A full SQL schema is included with the workflow download.
Output
What each post includes
Every post written by the agent is a complete, production-ready article. Nothing is left as a stub or placeholder — the workflow either produces a full output or fails visibly.
Full HTML article
4,000+ words, structured headings, lists, tables
SEO metadata
60-char title · 155-char meta · primary + 8 secondary keywords
Featured image
AI-generated 16:9 JPEG via Flux Pro, no text/logos
2 editorial section images
Contextual in-article images via Flux Dev
Citations & references
8+ in-text citations with live URLs, numbered reference list
FAQ section
5 questions, 100–150 word answers each, schema-ready
Article schema markup
Article type with author, publisher, datePublished, image
Category + tags
AI-assigned category, topic tags, primary keyword slug
Adaptability
Built for any industry
The default configuration targets AI and technology topics. Swapping it to any other industry requires editing a single code node — the topic pool array and category seeds.
AI & Technology
DefaultThe default configuration. Covers AI agents, LLMs, developer tools, enterprise AI, and model updates. Ready to use out of the box.
- AI agent frameworks
- LLM benchmarks
- Developer tools
- Enterprise AI ROI
Legal & Compliance
Swap the topic pool for legal tech, contract AI, regulatory updates, and compliance automation. Point citations at legal publications.
- Contract analysis AI
- Regulatory compliance
- Legal research tools
- E-discovery automation
Finance & Banking
Configure topics around fintech, risk modelling, algorithmic trading, and fraud detection. The research step pulls from financial news sources.
- Fintech AI
- Risk modelling
- Fraud detection
- Wealth management AI
Healthcare & MedTech
Adapt the topic pool to clinical AI, drug discovery, radiology automation, and EHR optimisation. No medical advice guardrails are already included.
- Clinical decision support
- Drug discovery AI
- Radiology AI
- EHR automation
E-commerce & Retail
Cover product recommendation AI, inventory forecasting, personalisation engines, and customer service automation.
- Recommendation engines
- Demand forecasting
- Personalisation AI
- Returns automation
SaaS & Software
Target topics like AI feature integration, LLM-powered UX, developer productivity tools, and technical SEO for software products.
- AI feature development
- LLM integration
- Developer productivity
- Technical SEO
HR & Recruitment
Focus on talent acquisition AI, skills-based hiring, HR analytics, and workforce planning. Built for HR tech publishers.
- AI screening tools
- Skills assessment
- Workforce analytics
- Onboarding automation
Marketing Agencies
Produce client-ready thought leadership on content strategy, SEO automation, ad AI, and marketing attribution for any agency niche.
- Content strategy AI
- SEO automation
- Ad creative AI
- Marketing attribution
Setup
Five steps to go live
From zero to autonomous publishing. A self-hosted n8n instance is the only infrastructure you need to run it.
Import the workflow
Download the n8n workflow JSON from Twarx. In your n8n instance, go to Workflows → Import → paste or upload the JSON file.
Add your API credentials
Open each HTTP Request node and add your credentials: Anthropic API key, Serper API key, and FAL.ai API key. The Supabase nodes use your project URL and service role key.
Set up your Supabase schema
Run the included supabase-schema.sql in your Supabase SQL editor. This creates the posts, categories, authors, and post_images tables with correct indexes and RLS policies.
Customise your topic pool
Edit the "Generate 5 Topics" code node to replace or extend the topic pools. Change the 5 category names and their topic arrays to match your niche.
Activate and monitor
Toggle the workflow to Active. It will run every 3 minutes automatically. Monitor execution in n8n's execution log. Posts appear live in Supabase immediately.
Pricing
Own it outright
One payment, lifetime licence. The complete package runs in your own n8n instance with your own API keys — no subscription, no usage fees to us.
Content Automation
AI Blog Content Machine
- Complete n8n workflow JSON, ready to import
- Supabase schema SQL — posts, categories, authors, images
- Step-by-step setup guide for all four API integrations
- 30-minute live setup call with our team
- 12 months of workflow updates included
No payment — just your email
Enter your email and the JSON downloads instantly.
Get started
Ready to deploy it?
Download the complete n8n workflow JSON for free — just enter your email and import it into your own n8n. Need help getting it running? Reach out anytime.