8 Million TikTok Views From a Gaming PC Under a Desk
Oliver Henry had an old NVIDIA 2070 Super gaming PC collecting dust. Three months ago, he wiped it, installed Ubuntu, set up OpenClaw, and turned it into an autonomous AI agent named Larry.
Larry’s job was TikTok marketing — end to end. Research competitors. Generate content. Write hooks. Test formats. Create images. Post everything. Oliver wanted to touch nothing.
Within five days, Larry had generated 500,000 views. By the end of the first week, that number hit 8 million. The article Oliver wrote documenting the whole experiment pulled another 7 million views on its own. He released the Larry skill for free, and thousands of people tried to replicate the setup.
Everyone focused on the AI side — the prompts, the image generation, the content strategy. Almost nobody noticed the infrastructure that actually made it work.
Larry posts through Postiz. Every draft gets uploaded automatically. Every post is scheduled at the right time. Content gets cross-posted to Instagram, YouTube Shorts, and Threads simultaneously. Analytics are pulled back after every post so Larry knows which hooks drive views and which ones to drop. The feedback loop that made Larry smarter every single day ran through Postiz’s API.
8 million views wasn’t just good content. It was good content with an automated distribution system that never sleeps.
$73,000/Month Running 12 Apps at Age 21
Ernesto Lopez is 21 years old. He runs 12 mobile apps generating $73,000 per month in recurring revenue. He built most of those apps using AI coding tools — Claude Code, Cursor, OpenClaw. One person doing the work of an entire dev team.
But building apps is half the game. The other half is getting people to download them. Marketing across TikTok, Instagram, X, Reddit, YouTube, and Pinterest — for twelve different products — would normally require a team of at least five people.
Ernesto has no team. His OpenClaw agent Eddie handles the marketing. Eddie generates content for all 12 apps, adapts it for each platform, and ships everything through Postiz.
One command per post. Formatted correctly for each platform. Scheduled at optimal times. Performance data fed back so Eddie gets better at knowing what converts.
$73,000 a month. One person. Twelve apps. Zero manual posting.
Bootstrapped to $25K MRR With Zero Marketing Budget
Mau Baron follows the same pattern. Bootstrapped to $25,000 monthly recurring revenue building apps. No marketing budget. No agency. He wired up OpenClaw with the Larry skill, pointed it at Postiz, and let the system run.
Content generated. Distributed. Tracked. Optimized. All automatic. All flowing through Postiz’s scheduling and analytics pipeline.
Three different builders, three different niches, three different product categories — same distribution tool, same outsized results.
Why Content Distribution Is the Real Bottleneck in AI Marketing
AI has solved content creation. That part is done. Your agent can write tweets, TikTok scripts, LinkedIn posts, articles, and ad copy faster than any human team. The models are good enough. The tools are mature.
But nobody solved the other half until now.
You have an agent that generates 15 posts a day. Then what? You spend three hours manually copying them into different apps. Reformatting each one for each platform. Different aspect ratios. Different character limits. Different optimal posting times. Scheduling by hand. Checking five separate analytics dashboards.
Your agent does the creative work in minutes. You spend the rest of the day being a copy-paste machine.
That’s not automation. That’s a content factory with no delivery trucks.
How Postiz Works as an AI Agent’s Distribution Layer
Postiz is an open-source social media management platform that supports 30+ platforms: X, TikTok, LinkedIn, Instagram, Reddit, YouTube, Pinterest, Facebook, Threads, Bluesky, Mastodon, Discord, Slack, Medium, Dev.to, and more. What makes it different from tools like Buffer or Hootsuite is that it was built specifically for programmatic access — every feature is available through a CLI, REST API, and MCP (Model Context Protocol) server.
Here’s what that means in practice: your AI agent talks to Postiz the same way it talks to any other tool. Cleanly. Reliably. At 3 AM when nobody’s watching. No browser automation. No screen scraping. No fragile scripts that break when a button moves.
The Postiz CLI: One Command to 30+ Platforms
The Postiz CLI (npm install -g postiz) gives AI agents direct access to the full scheduling pipeline. Here’s what a typical agent workflow looks like:
# Authenticate (one-time OAuth2 device flow)
postiz auth:login
# Discover connected social accounts
postiz integrations:list
# Get platform-specific settings (character limits, formats, etc.)
postiz integrations:settings <integration-id>
# Upload media to Postiz CDN
postiz upload ./generated-image.png
# Schedule a post across multiple platforms
postiz posts:create \
-c "Your AI-generated content here" \
-m "https://uploads.postiz.com/image.jpg" \
-s "2026-04-20T10:00:00Z" \
-i "twitter-id,linkedin-id,tiktok-id"
# Pull analytics to feed back into the agent
postiz analytics:post <post-id>
Every command returns clean JSON that an agent can parse natively. The agent discovers what platforms are connected, understands each platform’s requirements, creates content, schedules it, and then pulls performance data to improve future posts — all without a human touching anything.
MCP Integration: AI Agents Talk to Postiz Natively
For AI agents built on frameworks that support the Model Context Protocol (MCP), Postiz exposes a full MCP server with eight standardized tools:
// Connect any MCP-compatible AI agent to Postiz
{
"mcpServers": {
"postiz": {
"url": "https://api.postiz.com/mcp/your-api-key"
}
}
}
The MCP tools include integrationList (discover connected accounts), integrationSchema (get platform rules and character limits), schedulePostTool (schedule, draft, or publish posts), generateImageTool (generate AI images), and generateVideoTool (create videos). An AI agent can go from “I need to post about this topic” to “content live on six platforms” without any human intervention.
REST API for Custom Integrations
For builders who want deeper control, the Postiz REST API (base URL: https://api.postiz.com/public/v1) exposes every feature programmatically:
# Create a post with platform-specific settings
curl -X POST https://api.postiz.com/public/v1/posts \
-H "Authorization: your-api-key" \
-H "Content-Type: application/json" \
-d '{
"type": "schedule",
"date": "2026-04-20T10:00:00.000Z",
"posts": [{
"integration": {"id": "abc123"},
"value": [{"content": "<p>Your content here</p>"}],
"settings": {"who_can_reply_post": "everyone"}
}]
}'
The API supports platform-specific settings for every channel — Reddit subreddit and flair selection, YouTube video privacy and tags, TikTok duet/stitch permissions, LinkedIn company pages, Instagram story vs. post types, and more. Each platform’s unique requirements are handled through a settings schema that agents can query dynamically.
The Agent Feedback Loop That Compounds Results
What makes this setup genuinely powerful isn’t just scheduling — it’s the feedback loop. Here’s the pattern that Oliver, Ernesto, and Mau all use:
- Generate content — The AI agent creates posts tailored for each platform’s format and audience.
- Schedule through Postiz — One API call distributes to all connected platforms with platform-specific formatting.
- Pull analytics — After posts go live, the agent queries Postiz for engagement data (views, likes, comments, shares).
- Optimize — The agent uses performance data to adjust its content strategy — which hooks work, which formats drive engagement, which posting times perform best.
- Repeat — The cycle runs continuously, getting smarter with each iteration.
This is why Larry went from zero to 8 million views in a week. It wasn’t a lucky viral post — it was a system that tested hundreds of variations, learned what worked, and doubled down automatically. The distribution infrastructure made the feedback loop possible.
Setting Up Your Own AI Marketing System With Postiz
If you want to replicate what these builders have done, here’s the technical setup in five steps:
Step 1: Create a Postiz Account and Connect Your Platforms
Sign up at postiz.com and connect your social media accounts. Postiz supports 30+ platforms, so connect everything you want to post to.
Step 2: Install the CLI and Authenticate
npm install -g postiz
postiz auth:login
postiz auth:status # Verify you're connected
Step 3: Discover Your Integration IDs
# List all connected accounts
postiz integrations:list
# Check what settings each platform needs
postiz integrations:settings <integration-id>
Step 4: Wire Up Your AI Agent
Whether you’re using OpenClaw, Claude, or a custom agent, point it at the Postiz CLI or MCP server. The agent needs to:
- Call
integrations:list to discover available platforms
- Call
integrations:settings to understand each platform’s requirements
- Generate content adapted to each platform’s format
- Call
posts:create with the right integration IDs and settings
- Call
analytics:post after publication to collect performance data
Step 5: Let It Run and Monitor
The key is to let the agent operate autonomously while monitoring results through Postiz’s analytics. The system compounds — each day’s data makes the next day’s content better. Oliver, Ernesto, and Mau all report spending 15 minutes or less per day reviewing what their agents produce.
Why This Matters Right Now
The gap between “person with a system” and “person without one” is already significant in 2026. It’s growing every month.
The best marketers this year won’t be the best writers or designers. They’ll be the best system builders — the people who wire up an AI content engine with automated distribution and let it compound week after week.
Consider the numbers from just these four builders:
| Builder |
Revenue |
Views |
Team Size |
Daily Time Spent |
| Oliver Henry |
— |
8M+ (one week) |
Solo |
~0 min |
| Ernesto Lopez |
$73,000/mo |
Millions |
Solo |
~15 min |
| Mau Baron |
$25,000 MRR |
Millions |
Solo |
~15 min |
| Jordy (WeeklyClaw) |
— |
4M+ (1.5 weeks) |
2-person team |
~15 min |
If you’re creating content with AI and still manually posting to multiple platforms, you’re leaving views — and revenue — on the table every single day.
Frequently Asked Questions
What is Postiz and how does it work with AI agents?
Postiz is an open-source social media management platform that supports scheduling and analytics across 30+ platforms. Unlike traditional scheduling tools, Postiz provides a CLI, REST API, and MCP server that allow AI agents to programmatically create, schedule, and analyze social media posts without any manual intervention.
How many social media platforms does Postiz support?
Postiz supports over 30 platforms including X (Twitter), TikTok, LinkedIn, Instagram, Reddit, YouTube, Pinterest, Facebook, Threads, Bluesky, Mastodon, Discord, Slack, Medium, Dev.to, Hashnode, WordPress, and more. Each platform’s unique requirements (character limits, media formats, posting rules) are handled through platform-specific settings.
Can I use Postiz with any AI agent framework?
Yes. Postiz offers three integration methods: a CLI tool (npm install -g postiz), a REST API, and an MCP (Model Context Protocol) server. The CLI and API return clean JSON, making them compatible with any AI agent framework. The MCP server provides native integration for MCP-compatible agents like Claude and OpenClaw.
Is Postiz free or open source?
Postiz is open source and offers both free and paid tiers. The open-source version can be self-hosted, and the cloud version at postiz.com provides managed hosting with additional features. The API is rate-limited to 30 requests per hour on standard plans.
How does the AI agent feedback loop work?
After an AI agent schedules posts through Postiz, it can query the analytics API to retrieve engagement metrics (views, likes, comments, shares) for each post. The agent uses this data to optimize future content — identifying which hooks, formats, and posting times perform best. This creates a compounding effect where content quality and engagement improve automatically over time.
Start Building Your AI Marketing System Today
The pattern is clear: AI agents handle content creation, Postiz handles distribution, and the feedback loop between them creates compounding growth that manual posting simply cannot match.
Oliver went from a dusty gaming PC to 8 million TikTok views in a week. Ernesto manages 12 apps’ marketing from a single command line. Mau bootstrapped to $25K MRR without spending a dollar on advertising.
The tools are available right now. Get started with Postiz for free and connect your first AI agent today. The difference between starting now and waiting another month is another month of compounding results you’ll never get back.
