When we launched our MCP server generator, it was our most viral and well received launch of all time.

Riding on that momentum, we built and launched mcpt.com, a registry for MCP servers, in record time. Then we shut it down less than a week later.

Many people called me to ask why, so I wanted to share this story about something that's rarely discussed in public: when and why products get sunsetted – even moderately successful ones.

Why we built mcpt.com

The Mintlify MCP generator had just become our most successful launch ever. With momentum building and confidence high, we saw an opportunity to solve a critical problem in the MCP ecosystem: discoverability.

Users were creating incredible MCP products, but they had no central place to showcase them. As a team that excels at solving visibility problems, a universal MCP marketplace seemed like the perfect next step.

From idea to launch

With two engineers by my side, we transformed our ambition into action. We dedicated an intense week to building the first version of this marketplace platform, fueled by our vision of connecting MCP creators with their audience.

The immediate traction

Launch day exceeded our wildest expectations. Within 24 hours, mcpt.com attracted over 3,000 unique visitors. Messages flooded in from developers both within and outside our customer base, all eager to submit their servers to get listed.

The validation was clear – there was significant demand for what we'd built.

The reality check and why we pulled the plug

But a few days after launch, reality started to set in. We realized three fatal truths:

• Building and supporting a marketplace demanded operational resources we simply didn't have as a lean team.

• Every new bet requires tradeoffs, and this one would have diverted focus from the core product thousands of companies rely on daily.

• Most importantly, we recognized that marketplace building isn't our core strength – we excel at creating developer tools for companies, not consumer marketplaces.

So just five days after launch, despite promising traction, we made the difficult decision to sunset mcpt.com.

The lesson: Stay focused

Looking back, it seems like an obvious miscalculation. But in the heat of building something new, it's remarkably easy to get swept up in excitement and temporarily lose sight of your north star.

We don't regret the experiment. It validated the appetite for MCP tooling and delivered a powerful reminder about the critical importance of staying focused, especially during periods of rapid growth.

Why share this story? 

While most "build in public" narratives are shared long after the fact with perfect hindsight, we believe there's value in sharing the messy middle – the experiments, the pivots, and the failures – while we're still figuring things out.