MCP servers break too. Now you’ll know why.

SAN FRANCISCO: #ai--Sentry, the application monitoring and error tracking platform used by over 4 million developers, today announced the launch of MCP Server Monitoring. It gives anyone building on top of the Model Context Protocol (MCP) a clearer view into what’s working (and what’s not) behind the scenes.

First introduced by Anthropic in November 2024, MCP has seen rapid adoption as more companies move to make their products AI-compatible and turn to MCP to give AI agents a safer/standardized way to use their tools and data. Gartner predicts that by 2026, 75% of API gateway vendors and 50% of iPaaS vendors will have MCP features. Despite this momentum, once an MCP server is live, its developers generally have limited visibility into how it’s working.

“We built this because we needed to debug problems in our own MCP server, and quickly learned they’re the same problems everyone building MCPs is having,” says Cody De Arkland, Head of Developer Experience at Sentry.

“Sentry’s MCP server helps developers debug issues as they build,” he adds. “It turns out people really want that. Soon after launch, our MCP shot past 30 million requests a month. That sort of scale inevitably brings new bugs of its own. Existing monitoring tools struggle with the context of what’s happening in an MCP server. We needed to know things like traffic load and AI client usage, which tools were getting called the most, which were slow or failing, and which inputs were causing things to break. We needed to know all of this without relying on users to tell us.”

With just a few lines of code, Sentry’s MCP monitoring helps dev teams answer questions like:

• Which clients are experiencing errors or using outdated transports in your MCP Server?
• Which tools are getting the most use?
• Which tools are running the slowest; which are erroring out?
• Why are more errors suddenly occurring? Did it start right as traffic spiked, or right after a new release went live?
• Are errors happening because of a change you made, or because a bot is hammering your server with malformed requests?
• Are errors only happening on one type of transport? Are HTTP clients timing out, while stdio is fine?

Sentry already helps 4 million developers and 140,000 orgs find and fix bugs in their apps fast. Now that same visibility extends to MCP servers, all in the same interface developers already use every day.

“MCP is the fastest-growing protocol of the AI era, but when an MCP server breaks it can be tough to figure out what went wrong,” says Sentry CEO Milin Desai. “Your app, your agents, your MCP — it’s all one flow. With the addition of MCP Server monitoring, Sentry gives developers the context they need across every layer, so they can find the bug anywhere in their application stack and get back to shipping.”

MCP monitoring is available today for anyone using Sentry’s JavaScript SDK, and can be up and running in minutes.

About Sentry

Sentry helps every developer detect, understand, and fix broken code, fast. Using Sentry’s debugging platform decreases resolution time from days to minutes, giving developers more time to do the stuff they love, all while making customers happier. Sentry is used by over 4 million developers and 140,000 organizations, including Disney+, Cloudflare, GitHub, Anthropic, Vercel, and Atlassian.

Fonte: Business Wire