GTM Engineering with Claude Code vs Agentic Web Publishing
// TL;DR
If you're a marketer or growth operator who wants to automate SEO, ads, content creation, and publishing end-to-end using AI agents, choose Cody Schneider's GTM Engineering with Claude Code. If you're a web developer or product owner who needs to make an existing website accessible to AI agents and in-browser copilots, choose Rachel Lee's Agentic Web Publishing Framework. These skills solve fundamentally different problems — one automates your go-to-market execution, the other redesigns how your web property is consumed by agents.
// HOW DO THEY COMPARE?
| Dimension | Cody Schneider GTM Engineering with Claude Code | Rachel Lee's Agentic Web Publishing Framework |
|---|---|---|
| Best For | Marketers and growth teams automating GTM execution (SEO, ads, content, outreach) | Web developers and product owners making sites agent-accessible |
| Primary Output | Published content, ad campaigns, keyword research, performance reports | MCP servers, MCP Apps, Web MCP tool registrations on existing pages |
| Complexity | Low — folder setup, API keys, plain-language prompts to Claude Code | Medium-High — requires knowledge of MCP spec, CSP, iframe sandboxing, transport layers |
| Time to First Result | Under 1 hour — first article or keyword report can ship same session | Several hours to days — requires auditing site structure, building tools, testing across clients |
| Prerequisites | Claude Code access, API keys for marketing tools (Keywords Everywhere, CMS, GSC), a terminal | Web development skills (HTML/CSS/JS), existing website or app, understanding of MCP protocol |
| Creator Background | Cody Schneider — growth marketer and serial entrepreneur focused on AI-driven GTM | Rachel Lee — web developer and standards advocate focused on agentic web architecture |
| Scalability Pattern | Loop the same workflow across hundreds of keywords or ad variations via parallel agents | One-time infrastructure build that serves unlimited agent clients from a single backend |
| Feedback Loop | Continuous — pulls GSC data back into Claude Code to optimize published assets | Indirect — relies on agent harness adoption and Web MCP browser support for feedback |
| Technical Depth Required | Low — designed for non-engineers who can describe tasks in plain language | High — requires fluency in web APIs, CSP configuration, MCP spec nuances |
What does GTM Engineering with Claude Code do?
Cody Schneider's GTM Engineering with Claude Code is a framework for delegating every repeatable go-to-market task to AI agents running inside Claude Code terminal sessions. You create a single project folder containing a `.env` file (all API keys) and a `CLAUDE.md` file (standing instructions), and from that point forward, every agent session launched from that folder can autonomously execute keyword research, content writing, CMS publishing, ad management, and performance analysis.
The core operating model is what Cody calls being the "Conductor" — you open multiple terminal windows running parallel Claude Code sessions, assign each one a different sub-task, and jockey between them. One agent scrapes Google's page-one results for a target keyword while another drafts a blog post using that research as source material. A third publishes the finished article to your CMS via API. The human's job is to have the idea, provide quality source material (including a personal voice transcript), and polish the final output. Everything in between — the "Middle Work" — belongs to the agent.
The framework also closes the loop: after content is live, you feed Google Search Console data back into Claude Code via Graph MCP to diagnose underperforming pages and generate optimization instructions. This Continuous Improvement Loop turns one-time content into compounding GTM assets.
What does Rachel Lee's Agentic Web Publishing Framework do?
Rachel Lee's framework solves a completely different problem: making an existing website simultaneously servable to three client types — humans in browsers, humans using AI agents (like Claude Desktop), and agents browsing autonomously (like in-browser copilots). It is an architectural pattern, not a marketing automation workflow.
The core idea is "One Server, Three Clients." You audit your site's navigation and content structure, then map each user action (list, search, filter, view detail) to an MCP tool. For agent harness users, you expose these tools via an HTTP MCP server endpoint. For in-browser agents, you register tools directly on your HTML pages using Web MCP — either declaratively with HTML attributes on forms or imperatively via `navigator.modelContext.registerTool()`.
The framework's most distinctive concept is the MCP App: a single self-contained HTML file (HTML + CSS + JS bundled together) returned by an MCP tool and rendered inline inside the agent harness. This replaces the "starfish design" chatbox pattern — where the user faces a blank prompt — with rich, interactive, visually guided experiences inside the agent itself. Rachel argues that chat-only interfaces are the CLI of agentic UX, and MCP Apps are the GUI.
How do they compare?
These two skills operate at entirely different layers of the AI agent stack.
GTM Engineering with Claude Code is an execution framework. It assumes your websites, CMS platforms, and ad tools already exist and focuses on automating the human labor of using them. It is marketing-centric, low-code, and optimized for speed-to-output. A non-technical marketer can be publishing AI-assisted content within an hour.
Agentic Web Publishing is an infrastructure framework. It assumes AI agents already exist and focuses on making your web properties ready to be consumed by them. It is developer-centric, technically demanding, and optimized for architectural durability. A developer may spend days properly implementing it, but the result serves every future agent client automatically.
The two are not competitors — they are complementary. GTM Engineering might use the very MCP tools that Agentic Web Publishing creates. A team could use Rachel's framework to make their site agent-accessible, then use Cody's framework to automate the marketing workflows that drive traffic to it.
Where they overlap is in the belief that AI agents are the present, not the future. Both reject waiting — Cody rejects waiting to automate your GTM, Rachel rejects waiting for users to come back to your website.
Which should you choose?
Choose GTM Engineering with Claude Code if you are a marketer, founder, or growth operator who wants to stop doing hands-on-keyboard execution work. If your day involves keyword research, writing blog posts, managing ad campaigns, publishing to a CMS, or analyzing performance dashboards, this framework will automate 80% of that work immediately. You do not need to be a developer. You need API keys and clear task descriptions.
Choose Agentic Web Publishing if you are a web developer or product owner responsible for a website, documentation site, media archive, or SaaS dashboard that needs to be usable inside AI agent environments. If your users are already working inside Claude, Copilot, or browser-based AI agents, this framework ensures your content reaches them there. You need web development skills and comfort with the MCP protocol.
Choose both if you are building a content-driven business where you need to both produce content at scale (GTM Engineering) and ensure that content is accessible to the agent-first audience (Agentic Web Publishing). The GTM side automates your creation and optimization pipeline; the Agentic Web side ensures the output is consumable by the next generation of clients.
For most readers landing on this comparison — marketers looking to ship more, faster — GTM Engineering with Claude Code is the clear starting point. It requires less technical skill, delivers results in hours instead of days, and directly impacts revenue-generating activities. Add Agentic Web Publishing when your web infrastructure strategy demands it.
// FREQUENTLY ASKED QUESTIONS
Can I use GTM Engineering with Claude Code if I'm not a developer?
Yes. GTM Engineering is explicitly designed for non-engineers. You describe tasks in plain language, and Claude Code handles the technical execution. The only setup is creating a project folder with API keys and a CLAUDE.md file — no coding required. Cody Schneider built the framework for marketers and growth operators, not developers.
Do I need to know MCP to use Rachel Lee's Agentic Web Publishing Framework?
Yes. The framework is built on the MCP protocol — you'll be creating MCP tools, configuring HTTP transport endpoints, building MCP Apps as bundled HTML files, and handling constraints like CSP and sandboxed iframes. Familiarity with web development (HTML, CSS, JavaScript) and the MCP spec is essential.
Can these two frameworks be used together?
Absolutely. They operate at different layers and complement each other. Use GTM Engineering to automate content creation, publishing, and optimization. Use Agentic Web Publishing to make that published content accessible to AI agents and in-browser copilots. A content-driven business benefits from both.
Which framework is faster to get started with?
GTM Engineering with Claude Code is significantly faster. You can set up the Stack-in-a-Folder, run your first keyword research, and publish an article within an hour. Agentic Web Publishing requires auditing your site structure, implementing MCP tools, building MCP Apps, and testing across three client types — expect hours to days.
What is the Stack-in-a-Folder pattern in GTM Engineering?
Stack-in-a-Folder is a single project directory containing a .env file with all your API keys and a CLAUDE.md file with standing agent instructions. Every Claude Code session launched from that folder automatically inherits access to your entire tool stack. Set it up once and reuse it for every campaign.
What is an MCP App and how is it different from a regular website?
An MCP App is a single self-contained HTML file (HTML + CSS + JS bundled) returned by an MCP tool and rendered inline inside an AI agent harness like Claude. Unlike a website, it runs in a sandboxed iframe with no local storage and no direct network access — all external actions go through the agent harness.
Does GTM Engineering only work for SEO and content marketing?
No. While SEO content is the most common example, the framework covers paid ads, cold outreach, customer experience, product feedback loops, and any go-to-market function with API-accessible tools. If a human used to click or type to get it done, GTM Engineering can automate it.
What does 'starfish design' mean in Rachel Lee's framework?
Starfish design is Rachel's term for the chatbox-only landing page pattern where users face a blank prompt with no visual cues or navigation. She compares it to the CLI before the GUI — functional but high-friction. MCP Apps replace this with rich, structured, visually guided interfaces inside the agent.