Moth MCP server

Moth is a small helper that plugs into your AI coding assistant. It takes an error message you paste in, studies it, and hands back a short, structured write-up of what probably went wrong and what to try first. It is handy if you keep hitting bugs in the same project and want your AI to remember what fixed them last time.

What this MCP server does

You paste an error into your AI and ask it to use Moth. Your AI sends the error text to Moth, which first blanks out anything that looks like a password or secret. Moth then tidies up the error, guesses which framework or tool produced it, and checks a small file in your project for fixes you have already tried. It hands back a short brief with a likely cause, a suggested first fix, and a note about fixes to avoid. Later, once you have actually tried a fix and run a check, your AI can tell Moth whether it passed or failed, and Moth saves that in your project.

Click to zoom

What you can do with it

• Analyze an error message and get a likely cause plus a suggested first fix
• See fixes that already worked or failed in this same project
• Get a short list of things to avoid based on past attempts
• Record whether a fix passed or failed after you actually ran a check
• Keep a small fix history inside your project folder
• Redact likely secrets from error text before it is analyzed or saved

Try asking your AI

• “Use Moth to analyze this error before fixing it: TypeError: Cannot read properties of undefined (reading 'map')”
• “Here is the build error, please run it through Moth and tell me the best first fix”
• “We tried the suggested fix and the test command passed, please record that with Moth”
• “That last fix did not work, the check failed, please save the failed result in Moth”

What it gives back to you

You get back a short, structured brief in the chat: a likely cause, a suggested first fix, a verification step, a confidence level, and any past fixes for this project. It also lists things to avoid. When you record a result, Moth just confirms it saved the outcome to your project's fix memory file. Nothing is edited in your code and no commands are run for you.

Before you start

Who it's for

Developers and technical teammates who debug the same project often and want their AI to keep a small, honest memory of what actually fixed things.