How is this different from connecting JumpCloud MCP directly?

Connected directly, an agent sees every tool the server exposes and holds the credential itself. Through TrustGate you choose which tools each agent gets, and every call is logged against the agent that made it.

What if the JumpCloud MCP API changes?

We track the JumpCloud MCP schema, so the tool catalogue stays current. Tools you have already allowed keep working, and anything new stays denied until you enable it.

Can I give different agents different JumpCloud tools?

Yes. Access is set per agent: one can get a handful of read tools, another the full set. You grant the whole MCP or single tools within it.

Where can I see what my agents did in JumpCloud?

Every call is logged against the agent that made it, with the tool it used. Activity can be sent on to your own monitoring.

Do I need to run the JumpCloud MCP server myself?

No. TrustGate connects to the JumpCloud MCP server and sits in front of it. Your agent only talks to your TrustGate URL, never to JumpCloud directly.

Connect JumpCloud MCP to your AI agents, securely.

Q: Is JumpCloud MCP secure?

The server itself exposes 117 tools, and some of them can write to your JumpCloud account. Behind TrustGate an agent only sees the tools you allow, requests and responses are inspected, and every call is logged. Anything outside the allowlist is denied before it reaches JumpCloud.

One secured endpoint for every JumpCloud call your AI agents make, scoped per user, inspected, and logged.

Get started for free

Get a demo

Try JumpCloud MCP now

What your agents can ask once JumpCloud MCP is connected.

Select the tools your AI agents can call

117 tools from the JumpCloud MCP schema.

Showing 117 tools

users_listreadLists the users in the organization.
user_getreadReturns a single user's details.
user_createhigh riskCreates a new user in the directory, which is a new identity other systems will trust.
user_deletehigh riskDeletes a user from the directory.
user_suspendhigh riskSuspends a user, removing their access across everything the directory governs.
user_reactivatehigh riskReactivates a suspended user, restoring their access.
user_unlockhigh riskUnlocks a locked user account.
user_force_password_resethigh riskForces a password reset on a user account.
user_send_reset_password_emailhigh riskSends a password reset email to a user, which is an outbound message to a real person.
user_send_activation_emailhigh riskSends an activation email to a user.
user_reset_mfahigh riskResets a user's multi-factor authentication enrolment.
user_add_mfahigh riskAdds multi-factor authentication to a user account.
user_remove_mfa_requirementhigh riskRemoves the MFA requirement from a user account, weakening the control every downstream system depends on.
admin_listreadLists the administrators in the organization.
admin_getreadReturns a single administrator's details.
provider_admins_listreadLists the provider administrators at the MSP level.
admin_updatehigh riskUpdates an administrator account.
admin_reset_totphigh riskResets an administrator's TOTP, removing their second factor.
admin_reactivatehigh riskReactivates an administrator account.
provider_admin_invitehigh riskInvites a provider administrator, granting reach across the managed organizations.
provider_admin_removehigh riskRemoves a provider administrator.
devices_listreadLists the managed devices in the organization.
device_getreadReturns a single device's details.
device_updatewriteUpdates a device record.
device_lockhigh riskSends a lock command to a managed device.
device_restarthigh riskRestarts a managed device, interrupting whatever the person was doing.
device_shutdownhigh riskShuts down a managed device.
device_erasehigh riskErases a managed device. Nothing undoes it.
bind_user_to_devicehigh riskBinds a user to a device, granting them the ability to sign in to it.
unbind_user_from_devicehigh riskUnbinds a user from a device, removing their sign-in access to it.
commands_listreadLists the commands configured in the organization.
command_getreadReturns a single command's definition.
command_results_listreadLists the results of command runs.
command_result_getreadReturns a single command result.
command_result_listreadLists command results for a given command.
command_devices_listreadLists the devices a command is associated with.
command_device_groups_listreadLists the device groups a command is associated with.
command_runhigh riskRuns a JumpCloud Command on the managed devices it targets, which amounts to arbitrary code execution across the fleet from a single tool call.
command_manage_associationshigh riskManages which devices and device groups a command is associated with, changing how far the next run reaches.
user_groups_listreadLists the user groups in the organization.
user_group_getreadReturns a single user group.
user_group_membershipreadReturns the membership of a user group.
device_groups_listreadLists the device groups in the organization.
device_group_getreadReturns a single device group.
device_group_membershipreadReturns the membership of a device group.
device_group_createwriteCreates a device group.
user_group_deletehigh riskDeletes a user group, removing whatever access its membership conferred.
device_group_deletehigh riskDeletes a device group.
manage_user_group_membershiphigh riskAdds or removes users from a user group, which is how access is granted and revoked in JumpCloud.
manage_device_group_membershiphigh riskAdds or removes devices from a device group, changing which policies and commands reach them.
applications_listreadLists the SSO applications configured in the organization.
application_getreadReturns a single SSO application.
application_associations_listreadLists which users and groups are associated with an application, which is effectively an access map.
application_catalog_listreadLists the applications available in the JumpCloud catalog.
application_import_schedules_listreadLists the import schedules configured for applications.
application_resources_listreadLists the resources belonging to an application.
application_resource_getreadReturns a single application resource.
saas_applications_listreadLists the SaaS applications tracked in the organization.
saas_application_getreadReturns a single SaaS application.
saas_licenses_listreadLists the SaaS licences held across the organization.
saas_application_licenses_listreadLists the licences for a specific SaaS application.
saas_application_accounts_listreadLists the accounts held in a specific SaaS application.
saas_account_getreadReturns a single SaaS account.
saas_application_createwriteCreates a SaaS application record.
saas_application_editwriteEdits a SaaS application record.
saas_application_reviewwriteSubmits a review for a SaaS application.
saas_application_deletehigh riskDeletes a SaaS application record.
saas_account_deletehigh riskDeletes a SaaS account, which lands in the third-party system rather than in JumpCloud.
workflows_listreadLists the workflows in the organization. Public preview.
workflow_getreadReturns a single workflow definition.
workflow_runs_listreadLists the runs of a workflow.
workflow_run_getreadReturns a single workflow run.
workflow_createwriteCreates a workflow, which is automation that keeps acting after the conversation ends.
workflow_updatewriteUpdates an existing workflow.
workflow_duplicatewriteDuplicates a workflow.
workflow_trigger_runhigh riskTriggers a workflow run immediately, executing whatever that workflow does.
workflow_deletehigh riskDeletes a workflow.
patch_coverage_getreadReturns patch coverage figures across the estate.
patch_summary_getreadReturns a summary of patch status.
patch_critical_devices_listreadLists the devices missing critical patches.
patch_critical_summary_getreadReturns a summary of critical patch status.
patch_kbs_listreadLists the knowledge-base articles involved in patching.
patch_device_groups_listreadLists the device groups in patch management scope.
patch_devices_listreadLists the devices in patch management scope.
identity_providers_listreadLists the identity providers configured in the organization.
identity_provider_getreadReturns a single identity provider's configuration.
idp_ldap_listreadLists the LDAP identity providers.
idp_ldap_getreadReturns a single LDAP identity provider.
idp_policies_listreadLists the policies attached to identity providers.
idp_policy_getreadReturns a single identity provider policy.
health_rules_listreadLists the health rules configured in the organization.
health_rule_getreadReturns a single health rule.
health_rules_stats_getreadReturns statistics across the health rules.
health_rule_templates_listreadLists the available health rule templates.
health_rule_template_getreadReturns a single health rule template.
alerts_listreadLists the alerts raised in the organization.
alerts_statsreadReturns statistics across the alerts.
alert_getreadReturns a single alert.
alert_occurrencesreadLists the occurrences of an alert.
active_directories_listreadLists the Active Directory instances connected to the organization.
active_directory_translation_rules_listreadLists the Active Directory translation rules.
active_directory_translation_rules_previewreadPreviews how a translation rule would apply, without committing it.
active_directory_translation_rules_recommendationreadReturns a recommendation for an Active Directory translation rule.
assets_listreadLists the assets tracked in the organization.
asset_getreadReturns a single asset record.
asset_updatewriteUpdates an asset record.
asset_deletehigh riskDeletes an asset record from the inventory.
access_requests_listreadLists the access requests raised in the organization.
access_request_getreadReturns a single access request.
policies_listreadLists the policies applied across the organization.
policy_getreadReturns a single policy.
applemdms_listreadLists the Apple MDM configurations.
applemdm_abm_devicesreadLists the devices enrolled through Apple Business Manager.
applemdm_os_versionsreadLists the Apple OS versions in use across the estate.
di_events_getreadRetrieves Directory Insights events, the organization's audit log of sign-ins and administrative actions.
softwareapp_listreadLists the software applications known to the organization.
search_api_executereadExecutes a natural-language search across the organization, proxying to the JumpCloud search execute endpoint.

What JumpCloud MCP exposes

JumpCloud MCP groups its 117 tools into 17 categories, each with a suggested starting policy.

All 117 tools as JSON, ready to paste.

Category	Covers	Access	Suggested policy
Users	Listing and reading users, creating and deleting them, suspending and reactivating, unlocking, forcing a password reset, sending reset-password and activation emails, and resetting, adding or removing the MFA requirement on an account	Read / Write	Eleven of these thirteen change someone's ability to sign in. user_reset_mfa, user_add_mfa and user_remove_mfa_requirement touch the control every downstream system relies on, so gate them individually rather than as a group
Groups	Listing and reading user and device groups with their memberships, creating and deleting device groups, deleting user groups, and managing membership of both	Read / Write	Group membership is how access is actually granted in JumpCloud, so manage_user_group_membership is an entitlement change wearing an administrative name. Deleting a group removes whatever access it conferred
SaaS	Listing and reading SaaS applications, licences and accounts, creating, editing and deleting an application, submitting an application review, and deleting a SaaS account	Read / Write	saas_account_delete removes a person's account in a third-party system rather than in JumpCloud, so its effect lands outside the directory. The licence and account reads are a clear picture of what the organization pays for and who uses it
Devices	Listing and reading devices, updating a device, locking, restarting and shutting one down, erasing it, and binding or unbinding a user to a device	Read / Write	device_erase wipes a machine and nothing undoes it. Lock, restart and shutdown take a person off their computer immediately, and binding changes who can sign in to it
Commands	Listing and reading commands and their results, running a command, listing the devices and device groups a command targets, and managing those associations	Read / Write	command_run executes a JumpCloud Command on managed devices, which is arbitrary code execution across the fleet from one tool call. command_manage_associations changes which devices a command reaches, so it widens the blast radius before the run
Workflows	Listing and reading workflows and their runs, creating, updating, duplicating and deleting a workflow, and triggering a run	Read / Write	In public preview. A workflow is automation that keeps acting after the conversation ends, and workflow_trigger_run fires it immediately, so creating and triggering deserve separate approval
Applications	Listing and reading SSO applications, their associations and resources, the application catalog and import schedules	Read	Read-only. application_associations_list shows which users and groups reach which SSO applications, which is an access map of the organization and worth treating as internal
Patch Management	Patch coverage and summary figures, critical devices and their summary, the knowledge-base articles involved, and the device groups and devices in scope	Read	Entirely read-only, and one of the safest areas here. The output is a list of unpatched machines, which is a vulnerability inventory, so treat the result as sensitive even though the call is not
Admins & Provider Admins	Listing and reading administrators, updating one, resetting an admin's TOTP, reactivating an admin, and listing, inviting or removing provider admins at the MSP level	Read / Write	admin_reset_totp removes an administrator's second factor and provider_admin_invite adds somebody with reach across managed organizations. These are privilege changes and belong behind a named human approval every time
Identity Providers	Listing and reading identity providers, LDAP identity providers and the policies attached to them	Read	Read-only. The configuration returned describes how authentication is federated, which is useful context for an agent and useful reconnaissance for anyone else
Health Monitoring	Listing and reading health rules, their statistics, and the rule templates available	Read	Read-only, with no ability to change or silence a rule through this server, which removes the usual risk of an agent quietly weakening monitoring
Alerts	Listing alerts with their statistics, reading one, and listing its occurrences	Read	Read-only and a reasonable default allow. Alerts cannot be acknowledged or dismissed here, so the agent can read the security picture but not change it
Active Directory	Listing Active Directory instances and their translation rules, previewing how a rule would apply, and getting a recommendation for one	Read	Read-only, and the preview and recommendation tools are the right shape: they show the effect of a rule change without committing it
Assets	Listing and reading assets, updating one, and deleting one	Read / Write	asset_delete removes the record rather than the hardware, but an asset inventory with gaps is a compliance problem of its own
Access Requests & Policies	Listing and reading access requests, and listing and reading the policies applied across the organization	Read	Read-only. Access requests show what people have asked for and policies show what is enforced, so together they describe the gap between intent and configuration
Apple MDM	Listing Apple MDM configurations, the devices enrolled through Apple Business Manager, and the OS versions in use	Read	Read-only inventory of the Apple estate. Pair it with the patch tools for a version picture rather than using the device write tools to act on what it returns
Search & Insights	Executing a natural-language search across the organization through the search proxy, listing software applications, and retrieving Directory Insights events	Read	di_events_get reaches the audit log itself, including sign-ins and administrative actions, which is the organization's security record. search_api_execute proxies a free-form query, so log the query text and not just the tool name

Connect JumpCloud MCP to your AI agent

Four steps, under ten minutes.

01
Sign in to NeuralTrust
Open the NeuralTrust platform, sign in and go to Getting started, then Connect your tools.
02
Choose your agent
When asked which agent you use, pick Claude Code, Codex, Cursor, Windsurf, Copilot or your own code.
03
Connect JumpCloud
Find JumpCloud in the tool list and connect it.
04
Connect your agent
Follow the steps on screen to add the config to your agent, then send the test prompt.

Permissions down to the single tool

You do not allow JumpCloud MCP or block it. You allow the exact tools each agent can call, one by one.

JumpCloud MCP	117 TOOLS AVAILABLE
Code review agent	4 tools, read only
Issue triage bot	9 tools
Platform team assistant	all 117 tools

Tool by toolConnecting the server is not the unit of access. Each tool is granted on its own.
Default denyA tool you have not enabled is never exposed to the agent.
Approval on writesTools that change or delete things can require a human to confirm.

Is JumpCloud MCP secure?

One endpoint instead of raw credentials. Every call is checked against policy and logged.

AI agentsConnect through your MCP client configuration.	NeuralTrustApply controls based on your configuration.	JumpCloud MCP117 documented tools across US, EU and IN endpoints, with an organizationId parameter for MSP child orgs. Tools only, no resources or prompts, and streamable HTTP only; stdio is not supported. An admin must enable the server per organization, API keys mirror the admin's exact permissions, OAuth sessions are individually revocable with a configurable auto-revoke window, and MCP-originated requests appear in Directory Insights.
Coding assistants	Connection & credentialsper user	Users13 tools
Internal copilots	Tool access policiesdeny by default	Groups11 tools
Workflow automations	Request & response inspectioninline	SaaS11 tools
Compatible MCP clients	Activity monitoringalerts and logs	All other tools82 tools