When is Vindaris a good fit for connecting an ai assistant over mcp?

Who it helps: Vindaris fits the PM, analyst or IC who already works in Claude, ChatGPT or Claude Code and wants to act on Vindaris from there, and the admin who decides whether to allow it. It suits scale-ups and mid-market companies, roughly 30 to 500 people, where people increasingly run their day through an AI client and want their strategy data in reach. Why it helps: As AI clients become where people draft and plan, bouncing out to a separate app to check a goal or add a task breaks the flow. Connecting an assistant over MCP lets it query and act on Vindaris within the user's own permissions, so the work happens where the person already is. Act on Vindaris from your AI client. A connected assistant can look up goals, create tasks and update KPIs over MCP, so you plan in the client you already use and the changes land in Vindaris. It works within your permissions, so the assistant can do only what you can. For example: A PM asks Claude to turn a planning note into three tasks under a goal, and they appear in Vindaris aligned to that goal without the PM switching apps. Keep the assistant inside its lane. The connection is scoped by the same permission groups as people, so an assistant acts on its user's work and nothing wider. Access stays consistent whether the person or their assistant makes the change. For example: An analyst's assistant can query and update their team's KPIs but cannot touch another team's strategy, because it inherits the analyst's group. Turn it on deliberately. AI Tools stays off until an admin enables it, so MCP access is a decision rather than a default. Tokens and a REST fallback cover clients that do not speak MCP, so access stays controlled. For example: An admin enables AI Tools for the product team only, issues a token, and leaves it off for the rest of the org until there is a reason to extend it. Vindaris maps your strategy, goals and KPIs down to every project and task, and the Work Graph flags execution risk wherever effort has drifted from the goal, with task-level context down to the Slack thread.

Help Center . Sync & integrations

Connecting an AI assistant over MCP

You want to ask Claude or ChatGPT about your Vindaris work, or have it create goals and tasks for you, from your own AI client.

In short

Turn on AI Tools, connect Claude, ChatGPT or Claude Code to Vindaris over MCP, and let the assistant query and act within the permissions you set.

Good fit if you want to ask Claude or ChatGPT about your Vindaris work, or have it create goals and tasks, from your own AI client.

Vindaris can expose your workspace to an external AI assistant over MCP (the Model Context Protocol), so you can ask Claude or ChatGPT about your goals, KPIs and tasks, or have it create and update them for you. The assistant acts as you, inside the same permissions and tenant isolation that govern you in the app.

MCP is off until an admin turns it on.

As an admin, turn MCP on in Admin → MCP.It is off by default. Choose which resources an assistant may reach and whether each is read-only or read-and-write, and which individual tools are allowed org-wide.
Open Settings → AI Tools (MCP) to connect your client.For Claude Desktop, ChatGPT or Codex, add the Vindaris server URL as a connector and sign in. For Claude Code, create a personal token and run the shown "claude mcp add" command with it.
Choose which tools your assistant may use.Toggle individual tools on for yourself. Anything the admin has disabled org-wide stays off regardless.
Ask your assistant to work with Vindaris.It can list and open goals, KPIs, projects, tasks and strategies, create and update them, generate a scorecard summary, fetch your Work Graph alerts, align work, post a comment that syncs back to the source tool, set a risk bell, or send an update reminder.

Why it works

Why is it off by default and so tightly scoped?

An assistant acting on your workspace should reach no further than you do. So MCP ships off, an admin opts the org in, and every call runs inside the same row-level tenant isolation and permission groups as the app, down to whether the assistant may touch only your own items, your team's, or others'. Each tool call is recorded in the audit log as an AI action, so there is always a trail of what the assistant did.

Note

Tokens and a REST fallback

Personal tokens can be revoked or set to expire from Settings → AI Tools, and each is rate-limited. If your client does not speak MCP, the same actions are available over a plain REST API with an OpenAPI spec, linked from the same tab.

When teams reach for this

MCP stays off until an admin turns on AI Tools. Once on, you connect Claude, ChatGPT or Claude Code to Vindaris over MCP, and the assistant can query and act within the permission group its user has, with tokens and a REST fallback for other clients.

Who it helps

Vindaris fits the PM, analyst or IC who already works in Claude, ChatGPT or Claude Code and wants to act on Vindaris from there, and the admin who decides whether to allow it. It suits scale-ups and mid-market companies, roughly 30 to 500 people, where people increasingly run their day through an AI client and want their strategy data in reach.

Why it helps

As AI clients become where people draft and plan, bouncing out to a separate app to check a goal or add a task breaks the flow. Connecting an assistant over MCP lets it query and act on Vindaris within the user's own permissions, so the work happens where the person already is.

In practice

Act on Vindaris from your AI client. A connected assistant can look up goals, create tasks and update KPIs over MCP, so you plan in the client you already use and the changes land in Vindaris. It works within your permissions, so the assistant can do only what you can.ExampleA PM asks Claude to turn a planning note into three tasks under a goal, and they appear in Vindaris aligned to that goal without the PM switching apps.
Keep the assistant inside its lane. The connection is scoped by the same permission groups as people, so an assistant acts on its user's work and nothing wider. Access stays consistent whether the person or their assistant makes the change.ExampleAn analyst's assistant can query and update their team's KPIs but cannot touch another team's strategy, because it inherits the analyst's group.
Turn it on deliberately. AI Tools stays off until an admin enables it, so MCP access is a decision rather than a default. Tokens and a REST fallback cover clients that do not speak MCP, so access stays controlled.ExampleAn admin enables AI Tools for the product team only, issues a token, and leaves it off for the rest of the org until there is a reason to extend it.

Vindaris maps your strategy, goals and KPIs down to every project and task, and the Work Graph flags execution risk wherever effort has drifted from the goal, with task-level context down to the Slack thread.

Try Vindaris free

Put this into practice

Vindaris keeps strategy, goals, projects and tasks in one place, and the Work Graph flags execution risk before it costs you. Start free, no credit card.