It's been a minute since my last PLG Recap. This edition brings you five LinkedIn posts at the intersection of AI agents, MCP implementations, and emerging patterns already transforming the landscape.

The pace of innovation in AI agents and Model Context Protocol (MCP) has accelerated dramatically, with breakthroughs reshaping how we build, connect, and think about product-led growth.

Let’s dive in!

But first, here is my recommended subscribe list 🔥 

1️⃣ Paweł Huryn's guide to creating MCP servers for AI without coding

Recap

• Paweł Huryn shares a no-code approach to creating MCP servers for AI, leveraging the workflow automation platform, n8n.

• This is huge for product teams looking to extend their AI capabilities without deep technical expertise.

• I tried this out personally and I am still playing around with it, but I truly believe these types of MCP use cases will consolidate a ton of marketing roles.

• With just 6 simple steps using n8n, you can:

  • Create a workflow with MCP Server trigger.

  • Connect to 100+ ready-to-use tools.

  • Publish your workflow and test it

• The real power? You can wrap existing product APIs so AI agents can use them, drastically simplifying AI integration.

The bigger picture:

• Your next users might not be human, they might be AI agents.

• Products that make it easy for AI agents to interact without deep developer effort will get pulled into more workflows, faster.

• Being agent-accessible will soon matter as much as being API-accessible. This is going to create a major distribution advantage!

• Products that embed into agent workflows early will expand faster and compound growth over time.

2️⃣ Keith Richman's breakdown of his $890/month AI tool stack

Recap:

• Keith Richman shares a transparent look at his $890/month investment across 15 different AI tools.

• His stack spans:

  • Foundation models for research and writing.

  • Video generation tools for ad and content creation.

  • Image generation for product visuals.

  • Web app builders and niche applications.

• His goal is to experiment with how small businesses can leverage AI to grow without heavy in-house investment.

• For teams building AI solutions, this is an invaluable window into how SMBs are actually adopting and evaluating these tools.

The bigger picture:

• Keith’s experiment highlights a key reality that right now AI isn’t consolidating workflows, it’s fragmenting them. Each capability now often demands a different tool, creating friction in the user journey.

• With users overwhelmed by fragmented workflows, the next opportunity is to consolidate multiple jobs into a single environment without losing modularity or flexibility.

3️⃣ Avi Chawla's 10 practical MCP, AI Agents, and RAG projects for engineers

Recap:

• Avi Chawla drops an incredibly actionable resource for product and engineering teams building with AI agents.

• His post breaks down 10 fully documented projects with open-source code:

  • MCP-powered Agentic RAG systems

  • Building local MCP servers

  • Multi-agent book writing workflows

  • RAG implementations with models like Llama 4 and DeepSeek

  • Specialized systems like Corrective RAG and RAG over audio

• Each project comes with full walkthroughs and GitHub repos, giving you a head start on shipping advanced AI features without starting from scratch.

The bigger picture:

• Open-source AI and modular system design are shrinking build times. Teams can now launch production-grade workflows in days instead of months, speeding up time-to-market and feature velocity.

• AI-native workflows are becoming the new baseline. Teams that move fast will deliver sharper user value, lower CAC, and pull ahead while others are still planning their roadmap

4️⃣ Amos Bar-Joseph's complete AI agent stack powering a $30M ARR business with just 3 founders

Recap:

• Amos Bar-Joseph reveals how his team is targeting $10M ARR per employee with a network of 20+ AI agents instead of traditional hiring.

• Their AI agent stack has delivered impressive results in just 30 days:

  • $1M+ in qualified pipeline generated

  • 25 customers onboarded with white-glove service

  • 20+ major features shipped

  • 500+ support tickets resolved

• Each founder operates with their own AI taskforce: Marketing/Sales agents (The Observer, The Hunter, The Gatekeeper) Product/CS agents (The Concierge, The Analyst, The Prototyper), Engineering agents (The Architect, The Auditor).

The bigger picture:

• AI-augmented teams are redefining PLG, scaling personalized experiences by automating execution and freeing people to focus where it matters most.

• The real opportunity ahead is designing systems that scale human impact, lower CAC, increase LTV, and deliver high-touch experiences without scaling operational overhead.

• This is more than a growth hack. It could be the blueprint for how the next generation of companies will grow.

5️⃣ Dharmesh Shah's analysis of Google's A2A (Agent-to-Agent) Protocol launch

Recap:

• Dharmesh Shah cuts through the hype around Google's new A2A (Agent-to-Agent) Protocol with an early-stage analysis.

• Key takeaways:

  • A2A is not replacing MCP. It is complementary.

  • It solves important needs like capability discovery, agent messaging, async task handling, and weaving human UX into agent workflows.

  • The protocol feels complex, which could slow adoption compared to lighter, more flexible protocols like MCP.

  • Early signs point to A2A being built for large enterprises and consulting firms, not for open, cross-organizational agent networks.

• Real-world adoption outside the Fortune 1000 remains an open question.

The bigger picture:

• The AI landscape is shifting from building individual smart agents to building systems that can coordinate, communicate, and work together.

• Open, modular protocols like MCP lower friction, speed up experimentation and create the conditions for early network effects.

• In multi-agent ecosystems, bottom-up adoption beats top-down standards.

• Protocols that make it easy for developers to build today will dominate enterprise adoption tomorrow.

TLDR: Key takeaways and why this all matters

AI and product growth are evolving fast, but not randomly. Clear patterns are starting to emerge.

Across all five posts, it’s clear that the products that rethink onboarding, activation, workflows, and expansion automation through the lens of AI agents will pull ahead faster than those that don't.

Here’s what stood out:

• AI agents are becoming a new class of users. Products that make it easy for agents to interact will get pulled into workflows faster.

• The big question = will users want one platform that does it all, or a flexible stack of tools that fit together easily? Either way, the ability to connect and integrate smoothly will shape which products stick.

• No-code is expanding into agent infrastructure. Lowering technical barriers will help products embed into AI-driven workflows faster.

• Simplicity matters. Lightweight, open standards like MCP are better positioned to drive early adoption than heavyweight, enterprise-focused protocols.

• PLG is evolving with AI. Products will need to serve not just human users, but agents and systems operating behind the scenes.

• The next wave of growth will belong to the products that make themselves indispensable to agents, systems, and workflows.

What about you?

What are you experimenting with? Are you exploring any AI agents, MCP projects, or new tools? Let me know!

I’m always curious about what others are building.

Meme of the day #12

Interested in starting a newsletter?

Beehiiv 🐝 is the best tool for newsletters. Free to start and they don’t take a cut of your earnings. Use my referral link for a 30-day trial + 20% OFF for 3 months.

Thanks for reading!

If you are interested in sponsoring this newsletter, simply reply!