Vanta MCP server

What the model can do

Vanta compliance automation: automated tests and their failing resources, controls and owners, vendors with risk levels and review dates, issues, vulnerabilities with SLA dates, evidence documents and the tenant's users. When you ask, can assign control and document owners, link evidence to a document, add, edit or delete a vendor, record and update vendor findings, add and edit risk scenarios, deactivate or reactivate a test item or vulnerabilities, and complete offboarding for named people.

Reads

• vanta_list_testsVanta's automated compliance tests: name, status (NEEDS_ATTENTION means failing), category, integrations, owner, last run…
• vanta_list_test_entitiesThe individual resources one Vanta test checks (an S3 bucket, an account, a laptop), failing ones by default, or the deactivated ones: name, type…
• vanta_list_controlsVanta controls: reference, name, description, source (Vanta or custom), domains, owner and last change.
• vanta_list_vendorsVanta's vendor register: name, website, services, status, category, inherent and residual risk, owner user ids, contract renewal…
• vanta_list_issuesVanta issues (audit findings, incidents, pen-test and scanner findings): reference, title, severity, status, type, source, due, detected…
• vanta_list_vulnerabilitiesVulnerabilities Vanta's scanners found: name, severity, CVSS, package, fixable and fixed version, remediate-by (SLA) date, first and last detected…
• vanta_list_documentsVanta evidence documents (policies, screenshots, reports a framework asks for): title, category, status (Needs document, Needs update, Not relevant…
• vanta_list_usersActive users of the Vanta tenant: id and name (no email).

Acts, when you ask

These change something in Vanta. 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.

• vanta_set_control_ownerMakes a Vanta user the owner of a control, replacing the current owner; the control then shows on that person's list in Vanta.
• vanta_set_document_ownerMakes a Vanta user the owner of an evidence document, replacing the current owner.
• vanta_add_document_linkAttaches a link as evidence to a Vanta document (for example a policy page or a ticket), which auditors with access to the document can open.
• vanta_create_vendorAdds a vendor to Vanta's vendor register with its name and any of website, services, status, category, risk levels, owners, contract dates and notes.
• vanta_update_vendorEdits a Vanta vendor; only the fields passed change.
• vanta_delete_vendorPermanently deletes a vendor from Vanta's register.
• vanta_add_vendor_findingRecords a finding against a Vanta vendor (what was found in its review) with a risk status: ACCEPT, REMEDIATE…
• vanta_update_vendor_findingChanges a vendor finding's text, risk status or remediation (for example closing the remediation).
• vanta_create_risk_scenarioAdds a risk scenario to Vanta's risk register with its description and any of likelihood and impact (1 to 5), residual scores, treatment…
• vanta_update_risk_scenarioEdits a Vanta risk scenario's description, scores, treatment, owner or note; only the fields passed change.
• vanta_deactivate_test_entityDeactivates one item a Vanta test checks (a bucket, account, laptop) with a reason, so it stops failing the test, until a date or indefinitely.
• vanta_reactivate_test_entityReactivates a deactivated test item so the test checks it again (and may fail on it).
• vanta_deactivate_vulnerabilitiesStops Vanta monitoring the named vulnerabilities (up to 50 ids) with a reason, until a date or indefinitely, optionally resuming when a fix appears.
• vanta_reactivate_vulnerabilitiesResumes Vanta monitoring of the named deactivated vulnerabilities (up to 50 ids).
• vanta_offboard_peopleCompletes Vanta offboarding for the named former employees (up to 20 person ids), recorded as acknowledged by acknowledger_id.

Connect it

First connect Vanta on Dashboard → Connections, then add Memtro to your client once. The Vanta 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 Vanta 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 tests are failing, and how do we fix them?
• › Which vulnerabilities pass their remediate-by date this month?
• › Which vendors have a security review due this quarter?
• › Which evidence documents need an update?
• › Make Dan Fletcher the owner of the access review control.

Questions

Is this an official Vanta MCP server?

No. It is Memtro's MCP server with Memtro's Vanta connector behind it: the connector reads Vanta through its API with the credentials you connect on Dashboard → Connections, and Vanta 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 Vanta?

Only when you ask, and only what it has been granted: assign or change a control's owner; assign or change a document's owner; link evidence (a URL) to a document; add a vendor; edit a vendor, including its status and risk levels; delete a vendor (permanent); record a vendor finding; update a vendor finding's risk status or remediation; add a risk scenario to the risk register; edit a risk scenario's scores, treatment or owner; deactivate a test item so it stops counting against a test; reactivate a deactivated test item; deactivate monitoring for named vulnerabilities; reactivate monitoring for named vulnerabilities; complete offboarding for named former employees (marks their unmonitored accounts deactivated). 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