Shipr Enters Beta: What We Have Learned Building an Agentic Coding Platform

Shipr is now in beta. We have been testing internally for the past month, and we are starting to expand access to external users.

Here is what we have learned so far.

Planning Is the Product

When we started building Shipr, we thought the magic would be in code generation. Better autocomplete, smarter suggestions, faster execution.

We were wrong.

The magic is in planning.

Most software projects do not fail because of bad code. They fail because of unclear requirements, missed edge cases, and scope that was not properly defined. The problems happen before anyone writes a line of code.

So we built a planning engine. Not a feature: the core of the product.

When you start a project in Shipr, you do not jump into coding. You have a conversation. What are you building? Who is it for? What are the key features? What is the simplest version that would be useful?

Shipr asks questions, surfaces decisions, and constructs a structured plan. By the time you start building, you know what you are building.

Visual Progress Matters More Than We Expected

We added a visual progress indicator almost as an afterthought. A simple display showing what is done, what is in progress, what is next.

It has become one of the most-loved features.

Building software is often a grind. You work for hours and it is hard to see what you have accomplished. The visual progress display changes that. You see the plan. You see yourself moving through it. Each completed step is visible, tangible progress.

This sounds trivial. It is not. The psychological effect of visible progress keeps people engaged through the tedious middle parts of projects.

VS Code Integration Was the Right Call

We debated whether to build Shipr as a standalone web app or integrate with existing tools. We decided to do both, but the VS Code extension has been the clear winner.

Developers have strong preferences about their environments. Asking them to switch to a new tool for planning and then back to their editor for coding creates friction. The VS Code extension removes that friction.

You stay in your editor. You access Shipr's planning and execution tools without context switching. The integration is tight enough that it feels native.

If you are building developer tools, meet developers where they are. Do not ask them to come to you.

The Waitlist Is Teaching Us About ICP

We have 25 people on the Shipr waitlist. That is not a huge number, but it is enough to start seeing patterns.

Who is most excited: - Technical founders building MVPs - Solo developers with multiple side projects - Agency developers shipping client work on tight timelines

Common pain points: - "I have more ideas than time to build them" - "The setup and configuration for each project is tedious" - "I know what I want to build but getting started is the hard part"

What they are not saying: Nobody mentions code generation as the main draw. Few people mention AI specifically. The appeal is speed and clarity, not technology. This is shaping how we talk about Shipr. Less "AI coding assistant," more "ship faster."

What Is Next

We are expanding the beta over the next few weeks.

We are also working on export to markdown (take your Shipr plan and use it in other tools), time estimation features, and collaboration so you can share plans with team members.

Longer term, we are exploring how Shipr fits with our sovereign infrastructure work. Imagine agentic coding tools that run entirely on infrastructure you control, with no code sent to third-party providers. For enterprises building proprietary software, this could unlock use cases that are not possible today.