Connect Vaizy to Claude or ChatGPT
Add Vaizy as a connector in Claude, ChatGPT, Microsoft Copilot Studio or another AI app that supports MCP connectors. Then ask about your projects in plain language, right from the AI app:
"Which milestones in Website relaunch are slipping, and why?"
"What is on the critical path, and what happens if the design phase slips two weeks?"
"Compare the current plan with the baseline."
"What if we add a designer to Website relaunch in March? What does it do to cost?"
"Which of my projects need attention first?"
"Show me the health of all my projects." (In Claude and ChatGPT this opens a dashboard; see Project health dashboard.)
"Log 6 hours on DL-12 for today." / "What did I log this week?"
"Add a risk to Website relaunch: the supplier may deliver late."
"Put Bea on the Website relaunch team and on MS-3."
"Comment on MS-2 that the client signed off the design."
"Create a Vaizy project from the plan in Office move.xlsx."
"Plan an office move for 40 people, starting in March, with a budget of 60,000 euros."
Connecting AI apps is available on the Pro and Enterprise plans. It is off until an owner or admin of the workspace turns it on (see For owners and admins).
Connect
The connector URL is:
https://api.vaizy.com/mcpYou also find it in your personal Settings > MCP.
Claude (claude.ai or Claude Desktop): open Settings > Connectors, choose Add custom connector, paste the URL and select Connect.
ChatGPT: open Settings > Apps & Connectors, turn on developer mode under Advanced, create a connector and paste the URL.
Claude Code: run claude mcp add --transport http vaizy https://api.vaizy.com/mcp, then /mcp to sign in.
Microsoft Copilot Studio: Vaizy has a certified Power Platform connector, so there is no URL to paste.
Open your agent and go to Tools > Add a tool > Model Context Protocol, then pick Vaizy MCP.
Create a connection and sign in with your Vaizy account.
Add the tool to the agent and make sure generative orchestration is on.
Your AI app then opens Vaizy in your browser:
Sign in to Vaizy, with your two-factor code if you use one.
Check where Vaizy sends you back to. Known AI apps such as claude.ai and chatgpt.com are marked Verified. If an unverified app uses the name of a well-known AI app, Vaizy shows a warning. Only continue with an unverified address if you started connecting it yourself.
Choose the workspace to connect. A connection works in one workspace only; connect again to use another.
Choose the access:
- Read only: the app can look at your projects, milestones, deliverables, risks, logged hours, team capacity and insights.
- Read and write: the app can also create and update projects, milestones, deliverables and risks, log and correct hours, put people on project and milestone teams, add virtual team members, add finance entries and status remarks, and comment on milestones and deliverables.
Select Allow access. You are sent back to the AI app.
What the app can do
A connected app acts as you, with your permissions, in the workspace you chose. It sees only what you can see in Vaizy: if you cannot see a project's finances, neither can the app. It never sees rates, salaries or your colleagues' email addresses.
The app can never delete anything. Deleting stays in Vaizy.
With Read and write access, your AI app asks you before every change. Changes show in the audit log under your name, marked with the app's name (for example "via Claude").
Putting someone on a milestone team works as it does in Vaizy: their team costs are generated for that milestone from their rate. Inviting new people to Vaizy and taking someone off a project stay in Vaizy.
Ready-made prompts
Vaizy adds a few ready-made requests to your AI app's prompt menu (in Claude, through the + menu; in Claude Code, as slash commands such as /mcp__vaizy__weekly_status_report). Each one gathers the right facts from Vaizy and writes them up:
Weekly status report: health, progress, what slipped, top risks and decisions needed, for one project or all of them. Optionally for a given audience, such as a steering committee.
Risk review: stale risks, risks that no longer match the schedule, and suggestions for new ones.
Budget check: spend against budget, the burn rate and the expected cost at completion.
Slip report: which milestones are slipping, why, and what they push out.
Import a project plan: brings a plan from a file or another tool into Vaizy (see below).
Whenever a prompt would change something, such as adding a risk or saving a status remark, the app shows you the change and waits for your go-ahead.
Project health dashboard
Ask for the health of one or more projects and Claude or ChatGPT shows a dashboard: each project on track, at risk or off track, with its schedule and cost performance (SPI and CPI), the forecast finish date, the expected cost at completion, and the milestones most likely to slip. Select a project name to open it in Vaizy.
A project is on track when both SPI and CPI are 0.95 or better and nothing is overdue, at risk from 0.85, and off track below that. Forecasts only appear for projects whose finances you can see. AI apps that can't show dashboards give the same figures as text.
Plan a project from a short brief
Describe the project in a few sentences and your AI app drafts the plan: the team, milestones, deliverables, risks and a planned budget. The team uses people already in your workspace, plus virtual team members with a role and rate for open roles.
The app shows you the plan with every rate it assumed marked, so you can correct them. After you agree, it creates the project in one step and Vaizy calculates the team costs from those rates. Virtual team members need no account, email, invitation or seat.
The app can also give milestones a colour, so phases stand out on the timeline.
Bring in a project from a spreadsheet or another tool
If your AI app can read the file (for example an Excel plan you attached, or one it reaches through its own SharePoint or OneDrive connection), ask it to create the project in Vaizy. The app reads the plan, shows you what it will create, and after you agree creates the project with its team, milestones, deliverables, risks and planned income and expenses in one step.
Milestones get MS-N codes, deliverables DL-N and risks RI-N, in the order of your sheet. Milestone dependencies ("Move can only start after Plan") come along too.
People in the plan are matched to your team by name or email. Someone who is not on your team yet can be added as a virtual team member with a rate, or left out.
If anything in the plan doesn't fit (a date Vaizy can't read, an unknown status), nothing is created, and the app tells you exactly what to fix.
If a project with the same name already exists, the app asks before creating a second one, so asking twice never makes a copy by accident.
One import holds up to 50 team rows, 100 milestones, 300 deliverables, 100 risks and 200 income and expense rows. For a bigger plan, the app adds the rest to the new project afterwards.
You need permission to create projects in the workspace.
Disconnect
Go to your personal Settings > MCP and disconnect the app. It loses access at once. You can also disconnect from the AI app itself.
A connection also stops working when you leave the workspace, when the workspace leaves the Pro and Enterprise plans, or when an owner or admin turns off AI connectors.
For owners and admins
In your workspace's Settings > MCP you can:
turn Allow AI connectors on or off for the workspace. It is off by default on every plan, so members can only connect once you turn it on. Turning it off stops every existing connection on its next request.
see every connection in the workspace, who made it and when it was last used, and disconnect any of them.
Connections are honoured in workspaces that require two-factor authentication or single sign-on, but only for their own workspace, and they can only be made from a sign-in that meets those requirements.
Troubleshooting
"None of your workspaces can connect AI apps right now." No workspace you belong to can connect yet. The workspace needs the Pro or Enterprise plan, and an owner or admin must turn on Allow AI connectors in the workspace's Settings > MCP.
"Connecting AI apps requires a Pro or Enterprise plan for this workspace." The workspace moved to another plan. Connections start working again once it is back on Pro or Enterprise.
"An owner or admin of this workspace has turned off AI connectors." Ask an owner or admin to turn Allow AI connectors back on. You don't need to reconnect.
"This workspace requires two-factor authentication. Set it up in Vaizy, then reconnect." Set up two-factor authentication in your personal Settings > Security, then disconnect and connect again.
"You are no longer an active member of this workspace." or "The Vaizy account behind this connection is no longer active." Your access to the workspace was turned off. Ask the workspace's owner or admin.
"This workspace is archived or no longer exists." Connect again and choose another workspace.
The app keeps asking you to sign in, or says the token is invalid, expired or revoked. The connection was disconnected, replaced by a newer one, or ended for security reasons. Remove Vaizy in your AI app and connect again.
The app says it can't make changes. The connection has Read only access. Connect again and choose Read and write.
The app can't find a project or doesn't show its finances. The app only sees what you can see in Vaizy, in the one workspace you chose. Check that you have access to the project in Vaizy, and that you connected the right workspace.
A list stops partway. To keep answers short, Vaizy sends a limited number of rows at a time. Ask for the next page, or narrow the question, for example to one project or a date range.
"Vaizy is not available right now." Something went wrong on our side. Try again in a few minutes. If it keeps happening, contact us through help.vaizy.com.
Your data
When you use a connected app, you send your Vaizy data to that AI app yourself. What the app does with it is covered by its own terms and privacy settings, not by Vaizy's. If your organisation has rules about which AI tools may see project data, ask your workspace admin before connecting.
Some things to know:
The AI app is not a Vaizy subprocessor. Your organisation chooses it, and it handles the data under its own terms. That is why an owner or admin has to turn AI connectors on: it is your organisation's decision that project data may leave Vaizy for that app.
Data may be processed outside the EU, depending on the AI app. Check where your AI vendor processes data, and use a company account (such as Claude for Work or ChatGPT Enterprise) rather than a personal one if your organisation requires it.
Deleting data in Vaizy does not delete it in the AI app. Anything the app already read stays in its chat history until you delete it there.
Only what's needed is shared. The app sees names and ids of colleagues, not their email addresses, rates or salaries.