UpGuard MCP server

What the model can do

UpGuard CyberRisk: your own security score and active risks, monitored vendors with scores, tiers, labels and assessment status, a vendor's risks and how they changed between dates, its security questionnaires, and threat-monitoring findings; when you ask, can start or stop monitoring a vendor, set its labels, tier or attributes, send it a questionnaire, comment on or change a questionnaire's status, and comment on, assign, close, reopen or request remediation of a threat.

Reads

• upguard_organisationYour own organisation in UpGuard: its security score and the score in each category (email security, encryption, network security…
• upguard_vendorsMonitored vendors with their score (0 to 950), tier, labels, lifecycle stage and assessment status.
• upguard_vendorOne monitored vendor in detail: score and category scores, risk counts by category and severity, domains, industry and the industry average, labels…
• upguard_vendor_risksA monitored vendor's active risks (findings such as an expired certificate or a missing DMARC record), with severity, category, remediation advice…
• upguard_vendor_risk_changesWhat changed in a vendor's risks between two dates: risks introduced and risks resolved, by group and highest severity.
• upguard_risksYour own organisation's active risks from UpGuard's external scanning, with severity, category, remediation advice and affected hostnames.
• upguard_vendor_questionnairesSecurity (or relationship) questionnaires sent to a vendor, oldest first, with status (sent, opened, returned, in review, complete, cancelled…
• upguard_threatsThreat Monitoring findings (dark web, open web and social media) with severity, status, types, source, dates and investigator; filter by module…

Acts, when you ask

These change something in UpGuard. The model is told to use them only when you ask for exactly that, and each one is granted by name: per team on a shared connection, by the owner on a personal one.

• upguard_monitor_vendorStarts monitoring a vendor in UpGuard, which uses one of the licence's monitored-vendor slots (it fails when all are used).
• upguard_unmonitor_vendorStops monitoring a vendor in UpGuard, removing it from the monitored vendors list (also a vendor kept without monitoring).
• upguard_set_vendor_labelsReplaces a monitored vendor's labels in UpGuard with exactly the labels given (labels not listed are removed).
• upguard_set_vendor_tierSets a monitored vendor's tier in UpGuard (0 removes the tier).
• upguard_update_vendor_attributesSets or removes custom attributes on a monitored vendor in UpGuard; a null value removes that attribute, others are left as they are.
• upguard_send_questionnaireSends a security questionnaire to the people you name at a vendor (outside the company): UpGuard emails each recipient an invitation to complete it…
• upguard_comment_questionnaireAdds a comment to a vendor questionnaire in UpGuard, on the whole questionnaire or on one question, as the colleague named by author_email.
• upguard_update_questionnaire_statusMoves a vendor questionnaire between review states in UpGuard: returned (awaiting review) and in review either way, or complete from either.
• upguard_comment_threatAdds a comment to a Threat Monitoring threat, seen by colleagues on the threat (private only applies to threats under remediation).
• upguard_assign_threat_investigatorAssigns a colleague as a threat's investigator in UpGuard and puts the threat into investigating.
• upguard_close_threatCloses a Threat Monitoring threat as a false positive, an accepted risk (justification required) or remediated (remediation_reason required…
• upguard_reopen_threatReopens a closed Threat Monitoring threat into investigating, with the named colleague as its investigator.
• upguard_request_threat_remediationCreates a remediation request for a threat in UpGuard, assigned to the email address given (UpGuard emails them the request with email_message)…

Connect it

First connect UpGuard on Dashboard → Connections, then add Memtro to your client once. The UpGuard tools appear alongside the tools of everything else you have connected; nothing is configured per connector.

claude plugin marketplace add memtro-org/claude-memtro
claude plugin install memtro@memtro --config api_key=sk-memtro-...

{
  "mcpServers": {
    "memtro": {
      "url": "https://platform.memtro.com/api/mcp",
      "headers": {
        "Authorization": "Bearer sk-memtro-..."
      }
    }
  }
}

{
  "mcp": {
    "memtro": {
      "type": "remote",
      "url": "https://platform.memtro.com/api/mcp",
      "headers": {
        "Authorization": "Bearer sk-memtro-..."
      }
    }
  }
}

Who may use it

The key identifies the person, and the server offers that person the tools they may use everywhere else in Memtro. A shared UpGuard connection is used by the teams an administrator granted it to, tool by tool, on the Permissions page; a personal connection is used by its owner alone, and its tools that act stay off until the owner allows them. See Teams and permissions and Connectors.

Ask it like this

• › Which monitored vendors have the lowest scores?
• › What risks did our payroll vendor introduce or resolve this month?
• › Which vendor questionnaires are past their due date?
• › What are our own critical risks right now?
• › Set the payroll vendor to tier 1.

Questions

Is this an official UpGuard MCP server?

No. It is Memtro's MCP server with Memtro's UpGuard connector behind it: the connector reads UpGuard through its API with the credentials you connect on Dashboard → Connections, and UpGuard publishes nothing here. The same endpoint carries every other system you connect and Memtro's memory tools.

What does it cost?

Nothing on top of Memtro. One price per organisation by its number of users, from $50 a month for up to 10 users after a 7-day free trial, with every connector, the MCP server, jobs and memory included. You bring your own model keys.

Can it change anything in UpGuard?

Only when you ask, and only what it has been granted: start monitoring a vendor (uses a licensed vendor slot); stop monitoring a vendor; replace a vendor's labels; set or clear a vendor's tier; set or remove a vendor's attributes; send a security questionnaire to named people at a vendor; comment on a vendor questionnaire; change a questionnaire's review status; comment on a threat; assign a threat's investigator; close a threat with a reason; reopen a closed threat; request remediation of a threat from a named person. Every other tool reads, searches or lists. An administrator grants each acting tool by name per team, and a personal connection acts only after its owner allows it.

Which clients work?

Any MCP client that speaks Streamable HTTP: Claude Code, Cursor, Claude Desktop, OpenCode, Codex, Zed, Windsurf, Cline and the Anthropic and OpenAI SDKs. There is no local process to run; the server is remote and authenticated with your key.

Other MCP servers · Engineering and operations

• GitHub MCP server
• AWS CloudWatch MCP server
• Sentry MCP server
• ServiceNow MCP server
• Okta MCP server
• Twilio MCP server