Docket MCP server

Docket is a shared to-do list for your coding projects. You can ask your AI assistant to add a quick note or a task, and it lands in one place instead of getting lost in the chat. It is handy if you use several AI coding tools and want one list that all of them can read and write to.

What this MCP server does

You tell your AI something like add a todo: buy milk. The AI passes that to the Docket helper running on your computer. Docket files the item under the project you are working in, based on the git remote of that folder, so you never type the project name. Later you can ask your AI what is open in this project, and it reads the list back to you. You can also open a small web dashboard at localhost:8787 to see and edit everything yourself.

Click to zoom

What you can do with it

• Add a quick todo or note without leaving your AI chat
• List what is open in the current project
• Mark an item as in progress so you do not duplicate work
• Edit the title, priority, due date, or link of an item
• Mark items as done when you finish them
• See the full change history of one item
• Delete items you no longer need

Try asking your AI

• “Add a todo: check the login error on staging”
• “What is open in this project right now?”
• “Claim the fix for the broken export button so nobody else picks it up”
• “Mark the buy milk todo as done”

What it gives back to you

You get short, plain answers in the chat. A list comes back as one line per item, scoped to the project you are in. When you add or change something, the AI confirms it and the item shows up on the dashboard at localhost:8787. Nothing fancy, just the items and their status.

Before you start

Who it's for

Developers and anyone who uses several AI coding tools and wants one shared list of small tasks across projects.