Productivity MCP servers

Docs, notes, calendars and task trackers.

Productivity MCP servers connect an assistant to the places your team's written work already lives — the wiki, the issue tracker, the notes, the docs. They are the difference between an agent that reasons about your project in the abstract and one that has read the spec.

What makes this category worth installing is retrieval, not automation. Most of the value shows up the first time you ask a question whose answer is buried in a document nobody has opened in six months, and the agent finds it, quotes it and links to it.

The write side is real too — creating issues, updating a page, filing a task — but it is the part worth introducing slowly. An agent that files three duplicate tickets is a nuisance your colleagues have to clean up.

Productivity MCP servers in this directory

What people build with productivity MCP servers

Answering from the wiki instead of from memory

Ask how a system is supposed to work and get the answer out of the document that defines it, with a link, rather than a plausible guess.

Turning a conversation into tracked work

Describe what needs doing and have the agent write it up as a properly formed issue in the right project with the right labels.

Catching up on a project

Ask what has changed on a board or in a space since last week and get a summary drawn from the actual items, not the notifications you skimmed.

Drafting into your own tools

Have the agent write a document where the document belongs, in the format your team reads it in, instead of in a chat window you then copy out of.

Choosing a productivity MCP server

  • Install the one holding your team's canonical written record first. Whichever tool the answers actually live in beats whichever tool has the nicer API.
  • Check how the server handles search. Retrieval quality across a large workspace is what separates these servers, far more than the length of their tool list.
  • Look at the permission model. Several of these tools issue integration tokens scoped to specific spaces or projects — use that, rather than a token that can see everything you can.
  • Start read-only if the workspace is shared. Anything the agent writes, your colleagues see.

Before you connect one

These servers reach documents written on the assumption that only your team would read them. Before connecting one to a shared workspace, check what the token can see, and confirm your organisation is comfortable with that content passing through an AI client.

Frequently asked questions

Will my whole workspace be sent to the model?

No. The agent calls search and read tools, and only what those calls return enters the conversation. In practice that is the handful of pages relevant to what you asked, not the workspace.

Can the agent edit documents my team depends on?

Only if the server exposes write tools and your token allows it. If that makes you uneasy — a reasonable instinct on a shared workspace — issue a read-only token and the question is settled.

Do these work with a personal account?

Yes. Every server in this category authenticates with a token you generate, so a personal workspace works exactly the same way as a company one.

Install any of these

Every server on this page carries a one-line install command on its own listing. Paste it into your client config and restart — the productivity tools appear in your next session.

Or keep browsing: all MCP server categories · the full MCP server list · Agent Skills.