How to Build an AI Content Engine with Claude Code

The System That Changed Everything

I went from spending 4 hours per video to mass producing 23 scripted videos in a single sitting.

Not hypothetical. Not "potential output." I sat down one morning, opened my terminal, and by lunch I had 23 videos with researched hooks, scripted talking points, and ready-to-film cards sitting in my pipeline. All I had to do was press record.

The system behind it is something I call a content engine. It lives inside Claude Code, which is an AI that runs on your computer, reads your files, and executes real commands. It's not a chatbot. It's closer to having a research assistant, content strategist, and scriptwriter sitting inside your laptop, waiting for instructions.

The engine handles five stages of content creation. Research, multiply, script, film, post. Each stage feeds the next. You control the creative decisions. The AI does the tedious work that used to eat your entire day.

This is the workflow:

Every step in that diagram is a single command. No switching between 8 tabs. No copy-pasting between tools. One system, one place, every day.

This guide walks you through building the entire thing from scratch. I'm assuming you've never touched any of this before. By the end, you'll have a working AI content engine that researches trends, generates video ideas, writes scripts in your voice, and publishes across every platform you use.

Part 1 gets Claude running on your computer in about 10 minutes. Part 2 turns it into a content engine. Let's build it.

If you get stuck at any point: Take a screenshot of whatever you're looking at and paste it into Claude, ChatGPT, or any LLM. Ask it to help you troubleshoot. This is how you should be working with AI now. Don't stare at an error message for 20 minutes. Screenshot it, paste it, get the answer, keep moving.

Step 1: Get Claude on Your Computer

Before we touch any content strategy, we need the brain of the operation running on your machine. This takes about 5 minutes.

What Claude Code Actually Is

You've probably used ChatGPT or Claude in a browser. You type a message, it responds, and that's it. Claude Code is different. It's an AI that lives on your computer, can read and write files, run commands, and connect to external tools. When you tell it to research a topic, it actually goes and scrapes real websites. When you tell it to write a script, it saves a real file to your computer. It does things. That's the difference.

Create Your Account

1. Go to claude.ai and create an account
2. Subscribe to Claude Pro ($20/month) - this gives you generous usage and is by far the simplest way to get started

If you prefer pay-as-you-go instead of a subscription, you can set up API billing at console.anthropic.com.

Install Claude Code

There are three ways to use Claude Code. Pick the one that fits you best:

Option A: Claude Desktop App (Recommended)

This is the easiest path. It's a real app with a visual interface - no terminal experience needed.

1. Download the Claude desktop app from claude.ai/download
2. Install it like any other app
3. Open it and sign in with your Anthropic account
4. Click the code toggle in the bottom left to enable Claude Code mode

That's it. You now have Claude Code running in a clean visual interface. When this guide shows terminal commands, you can type them directly into the Claude Code chat - it handles the terminal stuff for you.

Option B: VS Code or Cursor

If you already use a code editor, Claude Code works as an extension inside VS Code or Cursor. Install the "Claude Code" extension from the marketplace and sign in. Same capabilities, just lives inside your editor. If you don't know what VS Code is, skip this and use Option A.

Option C: Terminal (Power Users)

If you're comfortable with the command line, you can run Claude Code directly in your terminal:

npm install -g @anthropic-ai/claude-code

If you get an error saying npm is not found, you need Node.js first. Go to nodejs.org, download the LTS version, run the installer, then try again.

Once installed, run claude in your terminal to start it. Follow the authentication prompts to connect your account.

Whichever option you chose, try typing a message and make sure you get a response. If that works, you're ready for the next step.

Step 2: Download the Template Pack

Download the template pack. I already built every config file, skill file, and reference template you need. I tested all of it. It works. You can either spend hours creating these files from scratch, or you can enter your email, download the pack, and have everything ready in 30 seconds.

Here's what you're getting:

• CLAUDE.md template - Full configuration file with every section pre-formatted. Fill in your details and go.
• 7 skill files - daily-content-researcher, content-ideator, content-scripter, post-content, shortform-analysis, competitor-analyst, video-downloader. Each one tested and production-ready.
• 4 reference file templates - avatar.md, scripting-voice.md, viral-content-patterns.md, hook-swipe-file.md. Each with examples and fill-in-the-blank sections.
• Airtable CSV import - Creates your content pipeline with all the right columns. Import it and you're set.
• Supabase migration SQL - Run it once and all analytics tables are created.
• Setup checklist - A printable step-by-step list so you don't miss anything.

It's free. It costs you an email address. The rest of this guide assumes you have these files, so grab them now and keep moving.

Step 3: Set Up Your Project

Claude Code reads files from whatever folder you point it at. So the first thing we need is a dedicated folder for your content system, with one specific file that tells Claude who you are.

Create Your Folder

Create a new folder on your computer called content-engine. You can put it anywhere - your Desktop, Documents, wherever you'll remember.

If you're in the terminal, run:

mkdir -p ~/content-engine
cd ~/content-engine

If you downloaded the template pack: Unzip it into this folder. You'll see a CLAUDE.md file, a skills folder, and a reference folder already set up. Open CLAUDE.md and start filling in your details (your name, your platforms, your brand voice).

If you're building from scratch: Create a file called CLAUDE.md in this folder. This is the single most important file in your system.

What Is CLAUDE.md?

CLAUDE.md is a configuration file that Claude reads every time it starts a session. It tells the AI who you are, what platforms you post on, what your voice sounds like, and what rules to follow. Think of it as a briefing document. Without it, Claude is a general-purpose AI. With it, Claude is YOUR content strategist who knows your brand.

Here's the minimum you need to fill in:

# My Content Engine

## Business Config
- Business name: [Your name or brand]
- What you do: [One sentence about your content/business]
- Platforms: [List every platform you post on with handles]

## Brand Voice
- Tone: [3-4 adjectives that describe how you sound]
- Always: [What your content should feel like]
- Never: [What your content should NOT feel like]

## Niche Keywords
[List 10-15 keywords/topics you cover]

Be specific in the voice section. "Professional but friendly" is useless. "Talks like a 26-year-old who reads a lot and swears occasionally, mixes internet slang with business concepts, never sounds like a LinkedIn influencer" is useful. The more specific you are, the better Claude writes for you.

Step 4: Your First Conversation with Claude

Open Claude Code and point it at your content-engine folder. It reads your CLAUDE.md automatically. You'll see it load your config in the first few seconds.

Now try this. Type:

Give me 5 video ideas about [your niche]

Watch what comes back. Claude already knows your brand voice, your platforms, and your topic area from the CLAUDE.md you just wrote. The ideas it generates should sound like something you'd actually want to make.

This is Claude working without any external tools connected. It's already useful. But the content engine makes it dramatically more powerful by connecting it to real data sources, a pipeline to organize everything, and repeatable skills that run the same proven workflow every single time.

That's Part 1 done. Claude is running on your computer, it knows who you are, and it can already brainstorm with you. Part 2 turns it into a system.

Build the Content Engine

Part 1 gave you a smart AI assistant. Part 2 gives you a system. We're going to connect two tools (one at a time, only when we need them), install the skills that automate your workflow, and run a full content session end to end.

Step 5: Connect Your Research Tool

The first skill we're going to use scrapes trending topics from X, Reddit, GitHub Trending, and tech news every morning. To do that, Claude needs access to a research tool called Apify.

Apify is a web scraping platform with a free tier that's more than enough for daily research. Here's how to connect it:

1. Go to apify.com and create a free account
2. Go to Settings → Integrations and copy your API token
3. Add your token to your computer's environment so Claude can find it

On Mac/Linux, open your terminal and run:

echo 'export APIFY_TOKEN="your_token_here"' >> ~/.zshrc
source ~/.zshrc

On Windows, add APIFY_TOKEN as a system environment variable through Settings → System → Environment Variables.

Step 6: Connect Your Content Pipeline

As the engine finds ideas and writes scripts, it needs somewhere to store them. That's Airtable - think of it as a smart spreadsheet where every idea, script, and published post lives in one organized view.

1. Go to airtable.com and create a free account
2. Create a new base (Airtable's word for a project)
3. If you have the template pack: Import the CSV file. It pre-builds all the columns, views, and statuses you need.
4. If building from scratch: Create a table with these columns: Title, Status (Idea / Scripted / Filmed / Published), Platform, Hook, Script Notes, Date Created, Date Published.
5. Go to airtable.com/create/tokens and generate a personal access token with read/write permissions on your base

Connect it to Claude Code. In your terminal (or paste this into Claude's chat), run:

claude mcp add airtable -- npx -y @anthropic/claude-airtable-mcp --api-key YOUR_AIRTABLE_TOKEN

Replace YOUR_AIRTABLE_TOKEN with the actual token you generated. Once this runs, Claude can read your pipeline, create new entries, and update statuses - all from one place.

Update the ## Tools section in your CLAUDE.md with your Airtable base ID (you can find this in the Airtable URL when you open your base).

Step 7: Install the Skills

This is where the system comes to life.

Skills are reusable instruction sets that Claude follows when you call them. Think of each skill as a detailed job description for a specific task. When you type /daily-content-researcher, Claude doesn't wing it. It reads the skill file, follows the exact process, scrapes the sources you specified, applies the ranking criteria you set, and outputs results in a consistent format.

Without skills, Claude is smart but undirected. With skills, it runs your content operation the same way every single time.

The Content Engine Skills

/daily-content-researcher - Scrapes four sources every morning (X, Reddit, GitHub Trending, tech news), filters for your niche, and presents a ranked Top 10 list of what's worth making content about today. Each topic has a brief on why it's trending, how big the audience is, and a suggested angle. Your picks auto-save to Airtable.

/content-ideator - Takes one topic and goes deep. Researches what content already exists, identifies angles nobody has covered, and generates five distinct video variations. Each variation has a different hook angle, audience, and emotional driver. One topic becomes five videos.

/content-scripter - The workhorse. For each video, it generates six data-backed hook options ranked by estimated performance. You pick the one that hits hardest. Then it generates a filming card: hook, three talking points, a close, and a CTA. Not a word-for-word script (those sound dead on camera), but a tight structure you can riff from naturally.

/post-content - Publishes your edited video to every platform at once. Transcribes the video, generates platform-specific captions in your voice, and posts to Instagram, TikTok, YouTube, LinkedIn, X, and Threads in a single command. (Requires additional setup - covered in "Going Further" below.)

/shortform-analysis - Analyzes your published content to find patterns. Which hooks get the most views? Which formats drive engagement? What posting times work? Gives you plain-English recommendations. Run this weekly.

/competitor-analyst - Point it at any creator or brand and it breaks down their strategy. Posting frequency, hook patterns, format mix, top-performing content and why it worked.

/video-downloader - Downloads videos from any platform (YouTube, TikTok, Instagram, X) to your computer. One command, any URL.

How to Install

If you have the template pack: Copy the skills folder from the pack into ~/.claude/skills/. Done.

If building from scratch: Each skill lives in its own folder with a SKILL.md file that contains its instructions. Create the folder structure:

mkdir -p ~/.claude/skills/daily-content-researcher
mkdir -p ~/.claude/skills/content-ideator
mkdir -p ~/.claude/skills/content-scripter
mkdir -p ~/.claude/skills/post-content
mkdir -p ~/.claude/skills/shortform-analysis
mkdir -p ~/.claude/skills/competitor-analyst
mkdir -p ~/.claude/skills/video-downloader

Each folder needs a SKILL.md file with the full instructions for that skill. You can also write your own skills for anything - client workflows, email sequences, podcast prep. The format is the same: a markdown file that tells Claude exactly what to do, step by step.

You can also write your own skills for anything. Client workflows, email sequences, podcast prep. Same format: a markdown file with instructions. I add new ones all the time.

Step 8: Run Your First Full Session

Everything is installed. Apify is connected. Airtable is connected. Skills are loaded. Time to run the engine for real.

Open Claude Code in your content-engine folder. From here, the daily workflow follows five stages.

Stage 1: Research

Type:

/daily-content-researcher

Watch your screen come alive. Claude reaches out to X, Reddit, GitHub Trending, and tech news through Apify. It pulls back dozens of trending topics, filters for your niche keywords, and ranks them by three criteria: total addressable audience (how many people care?), demo-ability (can you show it on camera?), and hook potential (is there a surprising angle?).

After a minute or two, it presents a Top 10 list. Each entry has a title, why it's trending, the source, and a suggested content angle. Pick your 3-4 favorites. Claude saves them to your Airtable pipeline as Ideas.

Stage 2: Multiply

Type:

/content-ideator

Claude pulls your Ideas from Airtable and asks which ones to develop. For each topic, it does a deeper research pass, checks what content already exists, and identifies underserved angles.

Then it generates five video variations per topic. Not five copies. Five genuinely different approaches - a speed build demo, a reaction format, a tutorial, a hot take, a comparison. Each has a distinct hook, target audience, and thesis.

Started with 4 topics? You now have 20 potential videos. Pick the ones that excite you. Claude updates Airtable and moves them to "Ideated."

Stage 3: Script

Type:

/content-scripter

This is where it gets good. For each video, Claude generates six hook options ranked by estimated performance. You pick the one that hits hardest.

Then it builds a filming card: your hook, three talking points with specific examples, a close, and a CTA. Not a script you read word-for-word - a structure you internalize and riff from naturally on camera.

20 videos scripted? 20 ready-to-film cards sitting in Airtable. Status: "Scripted."

Stage 4: Film

This part is you. Press record, reference your filming card, and talk. The engine doesn't film for you (yet). But all the prep work that used to take hours is done. You know what you're saying, why you're saying it, and how to open strong.

Most people batch-film 5-10 videos in one session once the scripts are ready. That's a week or two of content, shot in one afternoon.

Stage 5: Post

Once your videos are edited, publishing is one command. The /post-content skill can push to Instagram Reels, TikTok, YouTube Shorts, LinkedIn, X, and Threads simultaneously - with platform-specific captions generated from a transcript of your video.

This requires a couple of additional tools (Dropbox and Late) that are covered in the "Going Further" section below. For now, post manually - you're still getting 90% of the engine's value from the research-to-script pipeline. The posting automation is the cherry on top.

That's the loop. Research, multiply, script, film, post. Every day. The AI does the 80% that's tedious. You do the 20% that requires your face, your voice, and your personality.

Connect Analytics

Once you've been posting for a week or two, you'll want to start tracking what works. That's where Supabase comes in - a free database that stores your content performance data and powers the /shortform-analysis skill.

1. Go to supabase.com and create a free account
2. Create a new project (pick any region, set a database password)
3. Open the SQL Editor in your Supabase dashboard
4. Run the migration SQL from the template pack (it creates all the tables for content metrics, hook performance, and engagement data)

With Supabase connected, every video you publish adds data. Every week you run /shortform-analysis, you get smarter insights. The hooks get sharper. The topic selection gets more accurate. Month one, the system is useful. Month three, it knows your brand better than most human assistants.

Set Up Multi-Platform Posting

This is the part that feels like cheating. One command, seven platforms, posted simultaneously. Here's what you need:

Dropbox (Video Storage)

The posting skill pulls videos from a Dropbox folder. Set up is quick:

1. Go to dropbox.com/developers/apps and create a new app (Scoped access → Full Dropbox)
2. Before generating a token, go to the Permissions tab and enable: files.metadata.read, files.content.read, files.content.write, sharing.write. Click Submit.
3. Go to the Settings tab and generate your access token
4. Add it to your environment: DROPBOX_ACCESS_TOKEN=your_token

Late (Multi-Platform Publishing)

Late connects your social accounts into one API so Claude can publish everywhere at once.

1. Go to getlate.dev and create a free account
2. Connect your social accounts (Instagram, TikTok, YouTube, LinkedIn, X, Threads - as many as you want)
3. Go to Settings → API and copy your API key
4. Add it to your environment: LATE_API_KEY=your_key

How It Works

Drop your edited video into Dropbox's /Content/ready/ folder. Type /post-content. Claude scans the folder, shows you what's queued, and asks which video to post. It downloads the video, transcribes it with Whisper, generates platform-specific captions in your voice, and publishes everywhere. After posting, it moves the file to /Content/posted/ so your queue stays clean.

From "I have an edited video" to "it's live on seven platforms" takes about 60 seconds.

Build Your Voice Profile

Skills tell Claude what to DO. Reference files tell Claude who you ARE. These files live in your project folder and get loaded whenever Claude writes content for you. They're optional - the engine works without them - but they make a noticeable difference in output quality.

avatar.md

Describes your audience in two tiers. Your MACRO audience (broad - people scrolling who might follow you) and your MICRO audience (specific - people who might buy from you or hire you). Age range, where they hang out, what they care about, what makes them stop scrolling.

scripting-voice.md

How you talk on camera. Not how you write - how you speak. Your catchphrases, words you overuse, energy level, verbal tics. Record yourself talking for 5 minutes, transcribe it (Claude can do this), and note what makes your speech patterns yours. The result: scripts that sound like you talking, not like an AI.

viral-content-patterns.md

Eight content archetypes that consistently go viral in short-form video. "Impossible demo," "versus comparison," "speed challenge," etc. Build this over time by noting patterns in your niche. Every time you see a video with outsized engagement, log the pattern.

hook-swipe-file.md

Proven hook templates organized by archetype. "I tested [tool] for 30 days and the results are insane." "Nobody is talking about this." The /content-scripter skill pulls from this file when generating hook options. The more hooks you add, the better the suggestions.

Create these in your project:

mkdir -p ~/content-engine/reference
touch ~/content-engine/reference/avatar.md
touch ~/content-engine/reference/scripting-voice.md
touch ~/content-engine/reference/viral-content-patterns.md
touch ~/content-engine/reference/hook-swipe-file.md

Start with avatar.md and scripting-voice.md. Those two alone make a real difference. Build the others over time as you collect data.

What This Looks Like in Practice

Numbers help, so here's what a real session produced.

I sat down on a Tuesday morning with zero ideas. Ran /daily-content-researcher and got 10 trending topics. Picked 8 of them. Ran /content-ideator and got 5 variations per topic, giving me 40 potential videos. Filtered that down to 23 that felt strong. Ran /content-scripter on all 23.

By noon I had 23 videos, each with a researched hook, a 3-beat filming card, and a CTA. All saved in Airtable. All organized by platform and format. I batch-filmed 8 of them that afternoon and scheduled them for the rest of the week.

That's not a special day. That's what happens when research, ideation, and scripting are automated. The bottleneck moves from "what do I post?" to "which of these 23 videos do I want to film first?" That's a much better problem to have.

The engine also improves over time. Every video you publish adds data. Every week you run /shortform-analysis, you get better insights into what works for YOUR specific audience. The hooks get sharper. The topic selection gets more accurate. The scripts get closer to your actual voice.

Month one, the system is useful. Month three, it knows your brand better than most human assistants would.

Follow Along

I post Claude Code tutorials and AI content creation walkthroughs daily. If you want to see the engine running live, watch me build new skills, or get updates when I add new features, follow along on Instagram.

If you build your own content engine using this guide, send me a screenshot. I want to see it. Seriously. Nothing is more interesting to me right now than watching other people take this system and make it their own.

The tools are free. The strategy is on this page. The only thing between you and a working AI content engine is about 30 minutes of setup.

Go build it.