Vanta

Fix failing tests and manage controls, policies, vendors and risk in Vanta from Claude.

Opens Agent Studio, where connecting is one click. The connector URL below works in any MCP client.

Paste it into any MCP client. Setup docs

What the Vanta connector does

The Vanta MCP server lets Claude read and manage a company's trust and compliance program in Vanta. Agents can surface failing tests and plan fixes, browse controls and framework requirements, and manage policies, vendors, risks and privacy records. Many of its 176 tools write, and each acts with the permissions of the Vanta user who signed in.

Verified connector

Listed by Anthropic as a partner connector in its Connectors Directory.

Connection checked by Agentman on .

Anthropic states this reflects the level of review a connector received, not a security audit.

Vanta tools (176)

  • addAnswerLibraryTags
  • addDocumentsToEvidenceRequest
  • addKnowledgeBaseResource
  • addKnowledgeBaseTags
  • addPendingAnswerLibraryQuestionAnswers
  • addTagsToCategory
  • approvePendingAnswerLibraryQuestionAnswers
  • assignTestOwnership
  • checkRemainingPolicyDrafts
  • createControllerProcessingActivity
  • createControls
  • createCustomDocuments
  • createDataController
  • createDataInventoryCustomFields
  • createDataSubjectRequest
  • createPolicies
  • createPolicyControlMappings
  • createPrivacyAssessment
  • createPrivacyAssessmentVersion
  • createProcessorCategoryOfProcessing
  • createRiskScenarios
  • createTPRMVendorCustomFields
  • createTagCategories
  • createThirdPartyTicket
  • createTprmCustomResourceTypes
  • createVendors
  • createVendorsFromDiscovered
  • deactivateControls
  • deactivateTestEntities
  • deactivateTests
  • deleteControllerProcessingActivity
  • deleteDataController
  • deleteDataInventoryCustomField
  • deletePolicy
  • deletePrivacyAssessmentSeries
  • deleteProcessorCategoryOfProcessing
  • deleteTagCategories
  • downloadPolicy
  • editAnswerLibraryQuestionAnswers
  • editPendingAnswerLibraryQuestionAnswers
  • evaluateEvidence
  • fetchTrustCenterContent
  • generatePolicy
  • getAccessReviewDetails
  • getAgentRemediationPrompt
  • getAnswerLibraryEntriesCount
  • getAnswerLibraryQuestionAnswers
  • getAnswerLibraryTags
  • getAssessmentFindings
  • getAutomatedTestDetails
  • getControlDetailsAndStatus
  • getControlToPolicyMappingSuggestions
  • getControllerProcessingActivityDetails
  • getCreateThirdPartyTicketOptions
  • getDataSubjectRequestDetails
  • getEnabledRelationshipOptions
  • getEvidenceRequestDetailsTool
  • getFrameworkRequirements
  • getFrameworkSections
  • getKnowledgeBaseInsights
  • getKnowledgeBaseTags
  • getPendingAnswerLibraryQuestionAnswers
  • getPersonGroups
  • getPoliciesAcceptanceStats
  • getPolicyContents
  • getPolicyContradictions
  • getPolicyTemplateHtml
  • getPrivacyAssessmentDetails
  • getProcessorCategoryOfProcessingDetails
  • getQuestionnaireDetails
  • getRiskScenario
  • getRiskSettings
  • getSlaInconsistencies
  • getSlas
  • getTprmAssessmentTypeRules
  • getUploadedEvidenceContents
  • getVendorAssessmentMetadata
  • linkControllerProcessingActivityAndAssessmentSeries
  • linkObjectsToRiskScenario
  • linkRiskScenarioToImpactAssessment
  • linkThirdPartyTicketToObject
  • listAccessAccounts
  • listAccessAccountsForUser
  • listAccessRequests
  • listAccessReviews
  • listAccessSystems
  • listAssessmentQuestionsWithResponses
  • listAssessments
  • listAuditRequests
  • listAudits
  • listCommitmentTypes
  • listCommitments
  • listComputers
  • listConnectedIntegrations
  • listContracts
  • listControllerProcessingActivities
  • listControls
  • listDataControllers
  • listDataInventoryCustomFields
  • listDataProcessingVendors
  • listDataSubjectRequests
  • listDiscoveredVendors
  • listEvidenceRequests
  • listFrameworks
  • listIssues
  • listKnowledgeBaseResources
  • listOffboardingTasks
  • listPersonnel
  • listPersonnelGroups
  • listPolicies
  • listPolicyTemplates
  • listPolicyVersions
  • listPrivacyAssessmentTemplates
  • listPrivacyAssessments
  • listProcessorCategoriesOfProcessing
  • listQuestionnaires
  • listRemovedPolicies
  • listRiskRegisters
  • listRiskScenarios
  • listStandardCommitments
  • listStandardPapers
  • listTagCategories
  • listTests
  • listThirdPartyTicketingIntegrations
  • listTprmAssessmentAutomations
  • listTprmAssessmentTypes
  • listTprmCustomFields
  • listTprmCustomResourceTypes
  • listTprmVendorQuestionnaires
  • listTprmVendorSpreadsheetQuestionnaires
  • listVantaTaskObjectTypes
  • listVantaTasks
  • listVendors
  • listVulnerabilities
  • listVulnerableAssets
  • modifyTagCategories
  • modifyTags
  • reactivateControls
  • reactivateTestEntities
  • reactivateTests
  • rejectPendingAnswerLibraryQuestionAnswers
  • removeAnswerLibraryTags
  • removeDataSubjectRequestEvidence
  • removeKnowledgeBaseTags
  • removeTagsFromCategory
  • rerunQuestionnaire
  • searchKnowledgeBase
  • searchPolicyContents
  • searchVendorDocumentation
  • setTprmAssessmentAutomations
  • setTprmAssessmentTypeRules
  • submitImpactAssessmentForApproval
  • submitPolicyForApprovalDirectly
  • triggerTestRun
  • unlinkControllerProcessingActivityAndAssessmentSeries
  • unlinkRiskScenarioFromImpactAssessment
  • updateControl
  • updateControllerProcessingActivity
  • updateDataController
  • updateDataInventoryCustomField
  • updateDataSubjectRequest
  • updateIssue
  • updateKnowledgeBaseResource
  • updatePolicyDetails
  • updatePrivacyAssessment
  • updateProcessorCategoryOfProcessing
  • updateRiskCategory
  • updateRiskScenario
  • updateSla
  • updateVendors
  • uploadDataSubjectRequestEvidenceFromS3
  • uploadDocumentFromS3
  • uploadNewPolicyDraft
  • uploadPrivacyAssessmentFromS3
  • uploadPrivacyAssessmentVersionFromS3
  • upload_local_file

Tool names from Anthropic's directory listing. This server requires sign-in, so we could not read tool descriptions or parameter schemas.

Limits

  • Beta. Vanta labels the remote MCP server beta, released to all customers.
  • No read-only mode. An agent gets everything the connecting user's role allows.
  • No custom test authoring. Custom Tests can only be created in the Vanta dashboard; the MCP server can query them but not create or edit them, per Vanta.
  • Region-specific endpoints. The directory's URL serves US instances only.
  • Employee and personal data. Tools such as listPersonnel, listComputers, listAccessAccountsForUser and the data-subject-request tools return staff and individuals' details into your Claude conversation.
  • Tools vary by organization. Your tool list depends on enabled Vanta features and your role, so it may be shorter than 176.
  • Tool schemas are unverified. We could not read them without signing in, and called no tool.

Frequently asked questions

Which Vanta MCP URL should I use?

Use the URL for the region your Vanta instance runs in. Vanta hosts a separate MCP server per region: mcp.vanta.com for the United States, mcp.eu.vanta.com for Europe and mcp.aus.vanta.com for Australia, each at the /mcp path. Anthropic's directory listing points at the US server, so EU and Australian customers should add their regional URL as a custom connector instead.

Can I give an AI agent read-only access to Vanta?

No. Vanta's documentation says there is no read-only mode and no way to hand an agent a chosen subset of tools. A connection carries the full Vanta permissions of the user who authorized it, so scope access by choosing which account connects. Most MCP clients also let you switch off individual tools after connecting, which hides them from the assistant.

Why can't the agent upload files to Vanta?

Your AI application is probably blocking the temporary storage URL that Vanta uploads use. Vanta's troubleshooting guide says some desktop AI apps restrict which URLs an agent can reach, and the error message includes the blocked URL. Add that URL to the application's network allowlist, or ask your account owner or administrator to add it if you lack permission.

Sources

Use it in an agent

Put Vanta to work.

Connect Vanta once and your agents call these tools on their own: on a schedule, in a workflow, with nobody at the keyboard.

Call (650) 285-1019, Agentman.Our AI receptionist answers.