Member-only story
10 MCP Servers That Do Things I Didn’t Know MCP Could Do
I found some MCP servers that can give AI agents surprisingly powerful new abilities.
MCP servers have become one of the easiest ways to give AI coding tools access to external tools and data.
Most people know the usual ones.
GitHub. Filesystem. Playwright. Search. Databases.
But while looking through the MCP ecosystem, I found some servers that go way beyond those basic use cases.
Some can understand the structure of a codebase. Others can give an AI its own email inbox. One can even connect an AI agent to smart-home devices.
Here are 10 MCP servers that I think are worth knowing about.
1. Semble
Searching a large codebase with an AI can get expensive quickly.
The AI searches for something, opens several files, reads large chunks of code, and then tries to figure out what is actually relevant.
Semble takes a different approach.
It is a local semantic code-search engine that can be used through MCP. Instead of simply searching for exact words, it tries to find the parts of your code that are actually related to your question.
