MCP

The Simetrik MCP server puts Simetrik inside Claude directly. There's nothing to install and no terminal involved — you add one URL, sign in with your Simetrik account, and start asking.

It's the right path if you work in a chat window rather than a shell. If you'd rather your agent drove Simetrik from a terminal, use the CLI instead. Both reach the same platform, and you can use both.

Before you start

Two things must both hold, or the workspace won't appear when you connect:

• Your workspace has CLI/agent access enabled.
• Your role carries the Use CLI permission.

Both are granted by an account or security administrator. If you're missing either, send them Enable a role for CLI use — that page is the whole procedure.

Connect

https://mcp.us.simetrik.com/mcp

{
  "mcpServers": {
    "simetrik": {
      "url": "https://mcp.us.simetrik.com/mcp"
    }
  }
}

{
  "servers": {
    "simetrik": {
      "type": "http",
      "url": "https://mcp.us.simetrik.com/mcp"
    }
  }
}

What you can ask for

The first six run instantly. The last one goes to the Simetrik Agent — see below.

Read tools, session tools, and the Agent

The read tools hit read-only endpoints: there's no server-side path from any of them to a write. That's the guarantee, rather than the annotation — MCP tool annotations are advisory, and clients are told not to treat them as a security boundary.

Building work — new sources, reconciliations, rule sets, dashboards — goes through ask_simetrik_agent, which hands the task to the Simetrik Agent. Because it changes things, your client asks before every run. You confirm once to start it; collecting the answer afterwards is a read, so waiting never prompts you again.

What the MCP server does, and what your agent does

The MCP server isn't a one-click "ask anything, get a finished reconciliation" tool. It's a bridge, and knowing where the line falls saves a lot of frustration.

The server does:

• Return your real Simetrik configuration and data — resources, rules, rows, lineage.
• Enforce your permissions on every call. It can't reach anything your own user can't.
• Hand build-and-change work to the Simetrik Agent, which applies Simetrik's own methodology.

The server doesn't:

• Decide what's correct for your operation. It reports what's configured; whether that's the right control is your call.
• Upload local files. There's no filesystem here. Loading a file into a source needs the CLI or the web app.
• Guarantee the phrasing of an answer. Your client's model writes the prose. The server supplies the facts.

Multiple workspaces

Pick several at connect and name them in the question:

Your client routes each part to the right workspace. Ask "list my workspaces" to see which are active. Selections accumulate across reconnects, and you can add one mid-conversation without reconnecting — just ask.

Limits

• Agent runs are capped — 50 per day and 500 per month, per user. Read tools aren't capped. If you hit the cap, the server says so and tells you to ask your Simetrik contact to raise it.
• Agent runs take minutes, not seconds. A run that outlives the wait returns a session id, and your client collects the answer when it finishes.
• Sessions are long-lived. Sessions renew silently; you'll be asked to sign in again occasionally, and it's one click.
• US region only today. If your Simetrik account lives in another region, talk to your Simetrik contact.

Data handling

• Your permissions, never more. The server acts as you. A Viewer sees what a Viewer sees. Every call is authenticated, authorized and logged.
• Card numbers are truncated at the boundary. Where Simetrik identifies a card number in your data, the server sends only the first six and last four digits — the same masking the web app shows on screen — so a full card number isn't included in what reaches the model.
• Treat returned content as data, not instructions. Rows and descriptions in your workspace are text your agent reads. Review anything it proposes to change before approving it.
• Turning it off ends access. Remove it from your client and it's done — there's no key left behind to revoke.

Troubleshooting

No picker appeared. Expected when only one of your workspaces has access — you're already connected to it. Ask "list my workspaces" to confirm which one.

The workspace I want isn't on the picker. One of the two prerequisites is missing — the workspace gate or the Use CLI permission on your role. See Enable a role for CLI use.

"No workspaces available" after signing in. That screen means none of your workspaces has CLI/agent access yet — not that one is misconfigured. Same fix: an account or security administrator grants it.

A tool returns a permission error. Your role lacks that specific permission in that workspace. The same tool can work in one workspace and fail in another with the same account.

The Agent is taking minutes. That's expected for anything that plans or builds. Ask your client to check the reply rather than starting a second run — re-asking starts a new run and is billed as one.

I want a workspace I didn't pick at connect. Ask your client to add it. No reconnecting.

Next