Loading video
Keep AI Outreach Batches Small
John explains why smaller, personalized AI outreach batches are easier for Claude to handle.
Shorts for busy operators
Quick lessons on building, positioning, and automation. Each video gives you one idea you can use in your business today.
Watch on YouTubeLast updated: 1 August 2026
Latest lessons
Tap a cover to watch the full YouTube Short without leaving the page.
Keep the work personal, paced, and easy to review.
Loading video
John explains why smaller, personalized AI outreach batches are easier for Claude to handle.
Loading video
John verifies that an agent sent LinkedIn connection requests and will follow up after acceptance.
Loading video
John shows how a tighter rewrite makes an AI-drafted LinkedIn DM sound more direct and natural.
Loading video
John uses an Alex Hormozi skill to review contacts and connection notes before updating outbound messages.
Loading video
John explains how perfect, sequential timing can reveal automated LinkedIn activity.
Reach people without ignoring the limits of the platform.
Loading video
John shows how a connection-request note can replace paid InMail for first contact.
Loading video
John explains how Claude ranks leads by role, company, named tools, pain, and fit with the offer.
Loading video
John explains why AI outreach should read each Reddit community's rules before it acts.
Loading video
John explains why his LinkedIn tools restrict spam and respect platform limits.
Loading video
John explains why AI outreach should begin with human review instead of full automation.
Choose the right setup before asking Claude to act.
Loading video
John shows Claude adding a new competitor-versus-partner rule to a reusable skill.
Loading video
John explains why he prefers Claude Code for work that needs more tools and API access.
Loading video
John checks how a shared Claude skill keeps his API keys out and prompts each user to add their own.
Loading video
John explains how to move into a fresh Claude chat while carrying over the context that matters.
Loading video
John explains how Plan Mode works through the build, risks, options, and questions before Claude Code starts.
Move from a useful idea to a working, connected tool.
Loading video
John explains how MCP connections can fail while APIs are called only when needed.
Loading video
John explains why setup and ongoing operation call for different AI models.
Loading video
John explains why operators should launch useful products before they feel finished.
Loading video
John uses a toolbox comparison to explain how MCPs and APIs organize access to tools.
Loading video
John shows how to package a completed app build into a Claude skill that guides someone toward their own live app.
Make sharper choices about models, prompts, products, and buyers.
Loading video
John explains how model speed and intelligence affect which model fits a task.
Loading video
John explains how full project context and a clear problem help Claude find a better answer.
Loading video
John explains why a product for everyone converts no one.
Loading video
John explains why a useful ICP needs specific demographics, work context, and pains before choosing where to reach people.
Loading video
John argues that people who know how to direct AI gain value by producing apps and code faster.
An earlier prompt library for fast app ideas.
Loading video
The Short scrolls through more than 100 prompts for building instant apps with Lovable.
Ready to build
Bring the business process. We help you narrow the first version, connect the tools, and keep moving.
Join Agent-J+