Claude for Project Management: Projects, Connectors and a Shared Board

Claude is useful for project management in three places, each documented by Anthropic today. Projects give one project’s files and instructions a home, so every chat starts from the charter, the plan and the last status report instead of a blank page. Connectors let Claude read and act in the tools you already use, such as Slack, Google Drive or an issue tracker, with your own permissions. Cowork takes on longer, multi-step jobs and comes back with finished files. What Claude does not give you is shared state: nothing in Claude says, for the whole team, what is being worked on, by whom, and what is done. For that you connect a board, and with a remote MCP server that is a few clicks.

Claude Projects: a home for the project’s files and instructions

Anthropic’s help page What are projects?↗ (opens in a new tab) describes a project as a self-contained workspace with its own chat history and knowledge base. You upload documents to the project’s knowledge and write project instructions, and Claude uses both in every chat inside it. Projects are available on free accounts too, which are limited to five. The same page notes that a new version of projects is in beta, rolling out first to Claude Code users on Pro and Max; existing projects keep working as they are.

• Knowledge that earns its place: the project brief or charter, the current plan, the stakeholder list with who decides what, the last two or three status reports, and any contract or statement of work.
• Knowledge that goes stale: exported task lists and spreadsheets of who is doing what. They are out of date the day after you upload them.
• Instructions: how you want status written, who the audience is, and what Claude should ask before assuming.

You support the project manager for the Website Relaunch project.
Audience for status updates: the client's marketing director. Plain
English, no internal jargon, no more than 200 words.
Dates are US style (October 14, 2026).
Never invent a date, owner or number. If the knowledge does not say,
write "unknown" and list it under Questions.
When I paste meeting notes, sort them into decisions, actions (one
owner each) and open questions before doing anything else.

How to organize the knowledge and write instructions in depth is in Claude Projects best practices. Sharing a project with a team works on Team and Enterprise plans, but, as Anthropic’s page on project visibility and sharing↗ (opens in a new tab) says, your chats inside a shared project stay private unless you share them yourself. What that means for a team is covered in Claude Projects for teams.

Connectors: reaching the tools a PM already uses

According to Anthropic’s guide to using connectors↗ (opens in a new tab), connectors let Claude access your apps, retrieve data and take actions in them, for example creating an issue in Linear, sending a Slack message or searching Google Drive. Claude inherits each person’s permissions in the connected service, so it cannot reach a file or channel you could not. On Team and Enterprise plans an Owner enables a connector first and each person still signs in individually, and Owners can set each tool to always allowed, needing approval, or blocked.

For a project manager, the useful pattern is read first, write with approval: let Claude read the channel, the shared folder and the tracker freely, and require approval for anything that sends a message or changes a record. The three kinds of connector, and what each one can reach, are in Claude connectors explained.

Cowork: longer jobs that come back as files

Anthropic’s guide to getting started with Claude Cowork↗ (opens in a new tab) describes it as bringing Claude Code’s agentic capabilities to work beyond coding: you describe an outcome, step away, and come back to finished work such as formatted documents, spreadsheets with working formulas or slides. It is on paid plans, sessions can run in the cloud and keep going when you close your laptop, and scheduled tasks can run on a cadence you choose. Anthropic is rolling out a merged experience in which Chat and Cowork become one Claude, so on some accounts you simply ask in any conversation.

Good project management jobs for Cowork: turn a folder of meeting notes into a decisions and actions summary, build a stakeholder update deck from the last month of status reports, or reconcile a budget spreadsheet against invoices and list the differences. Fifteen worked examples with prompts are in Claude Cowork examples.

Drafting status updates and plans

Drafting is where Claude saves a project manager the most time, as long as the facts come from somewhere real. Give it the source, tell it the audience and the length, and forbid it from filling gaps.

Using the tasks and notes below, draft this week's status update
for the client.
- Three sections: Done this week, Next week, Needs a decision.
- Under 200 words. Plain English.
- Every item must come from the material below. If a date or owner
  is missing, say "not yet set" rather than guessing.
- End with a list of anything you left out and why.

[paste the board, the notes, or both]

The same approach drafts a project plan from a brief, a risk list from a plan, or an agenda from open questions. A library of sixteen of these, grouped by planning, triage, status, risks, meetings, retros and stakeholder updates, is in prompts for project management with AI.

Where Claude alone stops

• No shared state. A project holds files and instructions, not a live list of work. Nobody on the team can open Claude and see what is in progress.
• Private chats. What you and Claude worked out in your chat is not visible to a teammate using the same project unless you share it, and a shared chat is a snapshot.
• Status is only as fresh as the input. A draft built from last week’s export is last week’s status, written confidently.
• No record of who changed what. If Claude files an issue through a connector, the tracker records it; if it only suggests one in chat, nothing does.

None of this is a flaw in Claude. It is an assistant, not a system of record. The fix is to keep the work itself somewhere the whole team, and every assistant, reads and writes: a board.

Connecting fenbs to Claude over MCP

fenbs is a task board that AI assistants connect to over MCP as members, with a role and a name. In Claude it is a custom connector. Anthropic’s page on custom connectors using remote MCP↗ (opens in a new tab) says they work in Claude, Cowork and Claude Desktop, with free plans limited to one. On an individual Pro or Max plan:

1. Go to Customize, then Connectors, click “+” and choose “Add custom connector”.
2. Enter the server URL, https://fenbs.ai/api/mcp, and click Add.
3. Click Connect. fenbs opens in your browser: sign in, tick what the assistant may do (read, write, comment) and press Approve. Nothing is copied or pasted.
4. In a chat, switch the connector on from the “+” menu under Connectors, then ask Claude to read the board.

On Team and Enterprise plans an Owner adds it first under Organization settings, then Connectors, and each member connects it for themselves. In Claude Code the same server is one command, in the claude mcp add --transport http form that the Claude Code MCP documentation↗ (opens in a new tab) gives for remote servers; then run /mcp and choose fenbs to sign in through the browser.

claude mcp add --transport http fenbs https://fenbs.ai/api/mcp

Once connected, Claude starts with fenbs_whoami and fenbs_get_context, which gives it the team’s rules from the Decisions and rules page first, then the AI context notes. It files features, enhancements and bugs with fenbs_create_item, moves cards between To Do, Next Up, In Progress and Completed, and every change appears in the task’s history under its own name. A sign-in gets an access token that lasts an hour and refreshes; you can revoke the connection in Settings at any time. fenbs has no due dates, sprints or assignee field, so dates and owners go in the task note. The full walkthrough is in how to connect Claude to your project board with MCP.