🔒 "It's our own database, so it's safe" is a very natural thing to assume. With #WebMCP agents reading that data, it's worth a second look. WebMCP has a hint for this case, and JupyterLite is a nice example of using it.

Set untrustedContentHint: true on any tool that returns user-written text. That signals the host agent that the output came from an untrusted source, so the host can isolate it as data instead of treating it as instructions. JupyterLite sets the hint on all 19 tools that read notebook cells.

⚠️ A #WebMCP gotcha worth knowing about: throwing errors inside native registerTool() handlers. The browser hides your message and the agent only sees "DOMException: UnknownError". Return the error as a result instead, and the agent can fix its own call.

The #WebMCP hackathon winners are so much fun to dig through :D One pattern that shows up in several of them: tools that do the math in their own code (GPU, canvas, plain JS) and give the model a short, clear answer. Four examples, with links at the end.