Grafizum is a command center for developers coordinating AI coding agents across projects. It is one dashboard for projects, tasks, agent status, blockers, project memory and work waiting for review, and agents update it themselves over MCP.

Which AI coding agents does Grafizum support?

Any client that speaks remote MCP with OAuth: Claude Code, Cursor, Codex, Gemini CLI, Windsurf, Cline, Zed, Raycast, VS Code, Goose, opencode and LM Studio.

How do I connect an agent?

Add the server https://mcp.grafizum.com/mcp to your client. In Claude Code that is one command: claude mcp add --transport http grafizum https://mcp.grafizum.com/mcp. Then approve the connection once in the dashboard.

Do I need an API key?

No. The client signs in with OAuth 2.1 in the browser and renews its own token afterwards. There is no key to paste, so there is nothing to leak into your shell history.

Yes. The free plan is a plan, not a trial: two active projects, three subprojects each, 500 automated updates a month, no fixed limit on connected clients, no card.

What do the paid plans cost?

Starter is $9 per month (20 active projects, 10,000 automated updates a month). Team is $24 per month (4 people, 50 projects, 40,000 updates). Business is $99 per month (20 people, 200 projects, 200,000 updates). Additional people cost $5 per person per month. Annual billing gives two months free.

What counts as an automated update?

A successful write by an agent, for example creating or updating a task. Reads are never counted, so an agent can read the board as often as it needs.

What happens when I reach a limit?

The action is refused with the number that caused it and a way to raise it. Nothing is silently dropped.

What can a connected agent access?

Only the projects and scopes you granted to that connection. It can read projects, tasks, calendar and project memory, and write within its scopes. The secrets vault is never available over MCP, and you can revoke a connection at any time.

Can I see what an agent changed?

Yes. Every write lands on an append-only activity record with the name of the agent that made it.

Does Grafizum run my agents?

No. Agents keep running in your own terminal, editor or IDE. Grafizum is the board they read from and report to.

Do I need a team to use it?

No. It is built for one person running several projects at once. Plans with more people exist for when that changes.

Can different agents share one project?

Yes. All clients use the same endpoint, so Claude Code, Cursor and Codex can work from the same task queue, and each write shows which agent made it.

Can I stop an agent from writing on its own?

Yes. Switch the connection to manual mode. The agent still reads the board, but only writes when you ask in that turn.

Which languages are supported?

The website and dashboard are maintained in English, German and Italian.

FAQ

Frequently asked questions

Updated 10 November 2026

Short answers about Grafizum: what it is, which agents it supports, how setup works, what it costs and how access is limited. For setup details see the MCP setup guide; for every plan limit see Grafizum pricing.

Questions and answers

What is Grafizum?: Grafizum is a command center for developers coordinating AI coding agents across projects. It is one dashboard for projects, tasks, agent status, blockers, project memory and work waiting for review, and agents update it themselves over MCP.
Which AI coding agents does Grafizum support?: Any client that speaks remote MCP with OAuth: Claude Code, Cursor, Codex, Gemini CLI, Windsurf, Cline, Zed, Raycast, VS Code, Goose, opencode and LM Studio.
How do I connect an agent?: Add the server https://mcp.grafizum.com/mcp to your client. In Claude Code that is one command: claude mcp add --transport http grafizum https://mcp.grafizum.com/mcp. Then approve the connection once in the dashboard.
Do I need an API key?: No. The client signs in with OAuth 2.1 in the browser and renews its own token afterwards. There is no key to paste, so there is nothing to leak into your shell history.
Is Grafizum free?: Yes. The free plan is a plan, not a trial: two active projects, three subprojects each, 500 automated updates a month, no fixed limit on connected clients, no card.
What do the paid plans cost?: Starter is $9 per month (20 active projects, 10,000 automated updates a month). Team is $24 per month (4 people, 50 projects, 40,000 updates). Business is $99 per month (20 people, 200 projects, 200,000 updates). Additional people cost $5 per person per month. Annual billing gives two months free.
What counts as an automated update?: A successful write by an agent, for example creating or updating a task. Reads are never counted, so an agent can read the board as often as it needs.
What happens when I reach a limit?: The action is refused with the number that caused it and a way to raise it. Nothing is silently dropped.
What can a connected agent access?: Only the projects and scopes you granted to that connection. It can read projects, tasks, calendar and project memory, and write within its scopes. The secrets vault is never available over MCP, and you can revoke a connection at any time.
Can I see what an agent changed?: Yes. Every write lands on an append-only activity record with the name of the agent that made it.
Does Grafizum run my agents?: No. Agents keep running in your own terminal, editor or IDE. Grafizum is the board they read from and report to.
Do I need a team to use it?: No. It is built for one person running several projects at once. Plans with more people exist for when that changes.
Can different agents share one project?: Yes. All clients use the same endpoint, so Claude Code, Cursor and Codex can work from the same task queue, and each write shows which agent made it.
Can I stop an agent from writing on its own?: Yes. Switch the connection to manual mode. The agent still reads the board, but only writes when you ask in that turn.
Which languages are supported?: The website and dashboard are maintained in English, German and Italian.
How do I contact Grafizum?: Write to [email protected].

Frequently asked questions

Questions and answers

Read next