{
  "name": "Marketing MCP servers tracker",
  "description": "MCP servers from marketing automation and customer engagement vendors: how you connect, what an agent can read and change, whether any tool sends a message, and the safeguards each vendor documents.",
  "url": "https://fromhello.io/en/compare/marketing-mcp-servers",
  "publisher": "fromHello",
  "license": {
    "name": "CC BY 4.0",
    "url": "https://creativecommons.org/licenses/by/4.0/"
  },
  "credit": "fromHello, https://fromhello.io/en/compare/marketing-mcp-servers",
  "checked": "2026-10-12",
  "method": "Each row restates the vendor's own MCP documentation and tool list as read on the date in its checked field; it says what the vendor documents, not what we measured. fromHello's row describes our own product.",
  "fields": {
    "server": "official: published by the vendor (maker names another publisher); community: published by a third party",
    "status": "the vendor's own status label, or null",
    "connect": "remote-oauth: a hosted URL you sign in to; remote-api-key: a hosted URL that takes an API key or token; local: a process you run with a key",
    "access": "per area: change (create or edit), read (view only) or none (not documented); campaigns covers journeys, flows and workflows, templates covers message content",
    "sends": "yes: a documented tool sends or launches a message; opt-in: such a tool exists but stays off until you enable it; no: no documented tool sends a message"
  },
  "servers": [
    {
      "id": "customer-io",
      "vendor": "Customer.io",
      "server": "official",
      "maker": "Customer.io",
      "status": null,
      "connect": [
        "remote-oauth"
      ],
      "endpoints": [
        "https://mcp.customer.io/mcp",
        "https://mcp-eu.customer.io/mcp"
      ],
      "access": {
        "segments": "change",
        "campaigns": "change",
        "templates": "change",
        "contacts": "change",
        "analytics": "read"
      },
      "access_note": "Drafts unless an admin allows live edits.",
      "sends": "opt-in",
      "sends_note": "with the write:live scope, once an admin allows live edits",
      "safeguards": "A connection starts with the read scope; writing drafts, sending and configuring are separate scopes, an account admin must turn MCP on, and live data and sensitive attributes stay closed until an admin opens them. Tools support dry-run previews.",
      "watch_out": "Each user manages their own sessions: account admins can't view or revoke another user's MCP connection.",
      "summary": "Customer.io workspaces that want an agent across the whole API, with sending kept behind a scope of its own.",
      "capabilities": "Eight generic tools reach the full Journeys UI and Data Pipelines APIs: automations, segments, people, content and delivery metrics, with skills that guide multi-step tasks.",
      "checked": "2026-10-12",
      "source": "https://docs.customer.io/ai/mcp/get-started/",
      "more_sources": [],
      "publisher_row": false
    },
    {
      "id": "braze",
      "vendor": "Braze",
      "server": "official",
      "maker": "Braze",
      "status": null,
      "connect": [
        "remote-oauth"
      ],
      "endpoints": [
        "https://mcp.braze.com/mcp",
        "https://mcp.braze.eu/mcp"
      ],
      "access": {
        "segments": "read",
        "campaigns": "read",
        "templates": "change",
        "contacts": "none",
        "analytics": "read"
      },
      "access_note": "Campaign and segment writes need beta programs.",
      "sends": "opt-in",
      "sends_note": "launch_campaign, in the campaign APIs beta",
      "safeguards": "Access follows your Braze dashboard permissions and ends when your dashboard access does; no tool returns user-level PII, and Braze advises against your client's auto-mode.",
      "watch_out": "Workspaces that use IP allowlisting can't use the remote server, and Braze keeps an allowlist of supported clients, so others may be blocked at sign-in.",
      "summary": "Braze teams that want analytics and content work from an AI client, without exposing user-level data.",
      "capabilities": "Read and write tools for campaign and Canvas analytics, segments, custom attributes, email templates, content blocks, catalogs and media; Canvases are read-only.",
      "checked": "2026-10-12",
      "source": "https://www.braze.com/docs/user_guide/brazeai/mcp_server",
      "more_sources": [
        "https://www.braze.com/docs/user_guide/brazeai/mcp_server/available_api_functions/"
      ],
      "publisher_row": false
    },
    {
      "id": "iterable",
      "vendor": "Iterable",
      "server": "official",
      "maker": "Iterable",
      "status": "beta",
      "connect": [
        "local"
      ],
      "endpoints": [
        "@iterable/mcp (npx)"
      ],
      "access": {
        "segments": "change",
        "campaigns": "change",
        "templates": "change",
        "contacts": "change",
        "analytics": "read"
      },
      "access_note": "Writes and user PII are opt-in, key by key.",
      "sends": "opt-in",
      "sends_note": "a per-key sends flag, off by default",
      "safeguards": "Setup starts read-only, with no PII, writes or sends; each is a flag saved per API key, sends need writes, and keys can differ from one project to the next.",
      "watch_out": "In beta and local only, and creating a blast campaign schedules it for delivery: the API has no draft campaign.",
      "summary": "Iterable projects that want an agent to start read-only and gain writes, user data or sending one flag at a time.",
      "capabilities": "Tools map to Iterable's API: campaigns, templates, lists, catalogs, snippets, users and events, with journeys and campaigns that can be triggered once sends are on.",
      "checked": "2026-10-12",
      "source": "https://github.com/Iterable/mcp-server",
      "more_sources": [],
      "publisher_row": false
    },
    {
      "id": "klaviyo",
      "vendor": "Klaviyo",
      "server": "official",
      "maker": "Klaviyo",
      "status": null,
      "connect": [
        "remote-oauth",
        "local"
      ],
      "endpoints": [
        "https://mcp.klaviyo.com/mcp",
        "klaviyo-mcp-server (uvx)"
      ],
      "access": {
        "segments": "change",
        "campaigns": "change",
        "templates": "change",
        "contacts": "change",
        "analytics": "read"
      },
      "access_note": null,
      "sends": "yes",
      "sends_note": "send_campaign, to the campaign's audience",
      "safeguards": "read-only=true turns off every tool that writes, tools that read user-generated content can be switched off, and toolsets limits the server to the API scopes you choose. The remote server needs an Owner, Admin or Manager role.",
      "watch_out": "With writes on, send_campaign can send a campaign to its whole audience; for analysis, connect with read-only=true.",
      "summary": "Klaviyo accounts that want an agent across campaigns, flows, segments, profiles and reporting, with a read-only switch.",
      "capabilities": "Over 260 tools, from creating campaigns and flows to updating profiles and reading campaign and flow reports; a core-tools-only setting trims it to about 40 for clients with small context windows.",
      "checked": "2026-10-12",
      "source": "https://developers.klaviyo.com/en/docs/klaviyo_mcp_server",
      "more_sources": [
        "https://developers.klaviyo.com/en/docs/connect_to_the_klaviyo_mcp_server",
        "https://developers.klaviyo.com/en/docs/klaviyo_mcp_server_available_tools"
      ],
      "publisher_row": false
    },
    {
      "id": "hubspot",
      "vendor": "HubSpot",
      "server": "official",
      "maker": "HubSpot",
      "status": null,
      "connect": [
        "remote-oauth"
      ],
      "endpoints": [
        "https://mcp.hubspot.com"
      ],
      "access": {
        "segments": "change",
        "campaigns": "change",
        "templates": "change",
        "contacts": "change",
        "analytics": "read"
      },
      "access_note": "Email drafts only; workflows aren't listed.",
      "sends": "no",
      "sends_note": "marketing emails stay drafts",
      "safeguards": "Every action follows the user's HubSpot permissions and the scopes granted at install; when Sensitive Data is on, activities and conversations are blocked from the MCP server.",
      "watch_out": "Some writes go live at once: it can publish landing pages, website pages and blog posts.",
      "summary": "HubSpot accounts that want an agent on CRM records, campaigns and email drafts, within each user's own permissions.",
      "capabilities": "Reads CRM records, activities, campaigns, content and email analytics; creates and edits records, campaigns, static segments and marketing email drafts, and can publish landing pages, website pages and blog posts.",
      "checked": "2026-10-12",
      "source": "https://developers.hubspot.com/docs/apps/developer-platform/build-apps/integrate-with-the-remote-hubspot-mcp-server",
      "more_sources": [],
      "publisher_row": false
    },
    {
      "id": "activecampaign",
      "vendor": "ActiveCampaign",
      "server": "official",
      "maker": "ActiveCampaign",
      "status": null,
      "connect": [
        "remote-oauth"
      ],
      "endpoints": [
        "A private URL per account (Settings > Developer)"
      ],
      "access": {
        "segments": "change",
        "campaigns": "read",
        "templates": "change",
        "contacts": "change",
        "analytics": "read"
      },
      "access_note": "Reads campaigns and edits their message content.",
      "sends": "no",
      "sends_note": "it can add a contact to an automation",
      "safeguards": "Each account has its own private MCP URL, and connecting means signing in to ActiveCampaign and approving the connection.",
      "watch_out": "Adding a contact to an automation starts that automation's steps for them, and no read-only mode is documented.",
      "summary": "ActiveCampaign accounts that want an agent on contacts, lists, tags, deals and campaign content.",
      "capabilities": "Tools for contacts, tags, lists, custom fields, deals and pipelines, custom objects and email activity; it reads campaigns and edits a campaign's message content.",
      "checked": "2026-10-12",
      "source": "https://developers.activecampaign.com/page/mcp-available-tools",
      "more_sources": [
        "https://developers.activecampaign.com/page/connect-claude"
      ],
      "publisher_row": false
    },
    {
      "id": "brevo",
      "vendor": "Brevo",
      "server": "official",
      "maker": "Brevo",
      "status": null,
      "connect": [
        "remote-api-key"
      ],
      "endpoints": [
        "https://mcp.brevo.com/v1/brevo/mcp"
      ],
      "access": {
        "segments": "change",
        "campaigns": "change",
        "templates": "change",
        "contacts": "change",
        "analytics": "read"
      },
      "access_note": null,
      "sends": "yes",
      "sends_note": "SMS and WhatsApp campaigns",
      "safeguards": "You connect with an MCP token created among your API keys, and single-module endpoints, such as contacts only, give an agent a smaller set of tools.",
      "watch_out": "Its MCP page documents no read-only mode.",
      "summary": "Brevo accounts that want one connection across contacts, email, SMS and WhatsApp campaigns and the CRM.",
      "capabilities": "27 modules: contacts, lists, segments, email campaigns, campaign analytics, templates, SMS and WhatsApp campaigns, CRM deals and tasks, senders and domains, users and webhooks.",
      "checked": "2026-10-12",
      "source": "https://developers.brevo.com/docs/mcp-protocol",
      "more_sources": [],
      "publisher_row": false
    },
    {
      "id": "loops",
      "vendor": "Loops",
      "server": "official",
      "maker": "Loops",
      "status": null,
      "connect": [
        "remote-oauth"
      ],
      "endpoints": [
        "https://mcp.loops.so"
      ],
      "access": {
        "segments": "change",
        "campaigns": "change",
        "templates": "change",
        "contacts": "change",
        "analytics": "none"
      },
      "access_note": null,
      "sends": "yes",
      "sends_note": "published transactional emails",
      "safeguards": "Each client is authorized by signing in to Loops in your browser; the MCP page documents no other safeguard.",
      "watch_out": "Events can start published workflows, so an agent that sends events can trigger messages too.",
      "summary": "Loops teams that want an agent on contacts, lists, campaigns, workflows and transactional email.",
      "capabilities": "Finds, creates and updates contacts and their properties, manages mailing-list membership, sends events, creates and manages campaigns and workflows, and edits email content.",
      "checked": "2026-10-12",
      "source": "https://loops.so/docs/mcp-server",
      "more_sources": [],
      "publisher_row": false
    },
    {
      "id": "sequenzy",
      "vendor": "Sequenzy",
      "server": "official",
      "maker": "Sequenzy",
      "status": null,
      "connect": [
        "remote-oauth",
        "local"
      ],
      "endpoints": [
        "https://api.sequenzy.com/v1/mcp"
      ],
      "access": {
        "segments": "change",
        "campaigns": "change",
        "templates": "change",
        "contacts": "change",
        "analytics": "read"
      },
      "access_note": null,
      "sends": "opt-in",
      "sends_note": "a sender preset; the default preset can't deliver",
      "safeguards": "The default preset can draft but can't deliver, delete, change the team or create API keys; sending needs a sender preset or scopes, and the workspace role still applies.",
      "watch_out": "A removed permission can take up to five minutes to apply, while a widened one applies at once.",
      "summary": "Sequenzy workspaces that want an agent to draft campaigns and sequences, with delivery off by default.",
      "capabilities": "Covers subscribers, lists, tags, segments, campaigns, sequences, templates, transactional email, analytics, SMS and push, with presets that range from read-only to full access.",
      "checked": "2026-10-12",
      "source": "https://docs.sequenzy.com/concepts/mcp",
      "more_sources": [
        "https://docs.sequenzy.com/concepts/mcp-permissions"
      ],
      "publisher_row": false
    },
    {
      "id": "kit",
      "vendor": "Kit",
      "server": "official",
      "maker": "Kit",
      "status": null,
      "connect": [
        "remote-oauth"
      ],
      "endpoints": [
        "https://app.kit.com/mcp"
      ],
      "access": {
        "segments": "change",
        "campaigns": "change",
        "templates": "read",
        "contacts": "change",
        "analytics": "read"
      },
      "access_note": null,
      "sends": "yes",
      "sends_note": "broadcasts, with a confirmation",
      "safeguards": "Every tool carries an annotation, and Kit says sending broadcasts and deleting data always require explicit user confirmation; each token is limited to 120 requests a minute.",
      "watch_out": "Paid plans only: Creator and Creator Pro.",
      "summary": "Creators on a paid Kit plan who want an agent to manage subscribers and draft, schedule or send broadcasts.",
      "capabilities": "Maps Kit's v4 API one to one: subscribers, tags, segments, custom fields, sequences, broadcasts and stats, including composing, scheduling and sending broadcasts.",
      "checked": "2026-10-12",
      "source": "https://developers.kit.com/mcp/kit-mcp",
      "more_sources": [],
      "publisher_row": false
    },
    {
      "id": "resend",
      "vendor": "Resend",
      "server": "official",
      "maker": "Resend",
      "status": null,
      "connect": [
        "remote-oauth",
        "remote-api-key",
        "local"
      ],
      "endpoints": [
        "https://mcp.resend.com/mcp",
        "resend-mcp (npm)"
      ],
      "access": {
        "segments": "change",
        "campaigns": "change",
        "templates": "change",
        "contacts": "change",
        "analytics": "none"
      },
      "access_note": null,
      "sends": "yes",
      "sends_note": "emails and broadcasts",
      "safeguards": "The hosted server asks you to sign in and approve access, or takes an API key for headless clients; the open-source local server can limit which folders it reads attachments from.",
      "watch_out": "No read-only mode is documented, and the same connection can create API keys.",
      "summary": "Developers on Resend who want an agent to send, schedule and organize email across the whole platform.",
      "capabilities": "Sends and schedules emails, creates and sends broadcasts, builds automations, publishes templates, and manages contacts, segments, topics, domains, webhooks and API keys.",
      "checked": "2026-10-12",
      "source": "https://resend.com/docs/mcp-server",
      "more_sources": [],
      "publisher_row": false
    },
    {
      "id": "postmark",
      "vendor": "Postmark",
      "server": "official",
      "maker": "ActiveCampaign",
      "status": null,
      "connect": [
        "local"
      ],
      "endpoints": [
        "@activecampaign/postmark-mcp (npx)"
      ],
      "access": {
        "segments": "none",
        "campaigns": "none",
        "templates": "change",
        "contacts": "none",
        "analytics": "read"
      },
      "access_note": "Transactional email; no contacts or campaigns.",
      "sends": "yes",
      "sends_note": "single and batch sends",
      "safeguards": "Tools carry read-only and destructive annotations, so a client can run reads freely and ask before a send or a delete; the token limits it to one Postmark server.",
      "watch_out": "Local only, and one batch call can send up to 500 emails.",
      "summary": "Teams sending transactional email through Postmark who want an agent to send it, manage templates and diagnose delivery.",
      "capabilities": "24 tools: single sends and batches of up to 500 messages, template management, message search, bounces and suppressions, delivery stats and webhooks.",
      "checked": "2026-10-12",
      "source": "https://github.com/ActiveCampaign/postmark-mcp",
      "more_sources": [],
      "publisher_row": false
    },
    {
      "id": "mailchimp-transactional",
      "vendor": "Mailchimp Transactional",
      "server": "official",
      "maker": "Mailchimp Transactional",
      "status": null,
      "connect": [
        "remote-api-key"
      ],
      "endpoints": [
        "https://mandrillapp.com/mcp"
      ],
      "access": {
        "segments": "none",
        "campaigns": "none",
        "templates": "change",
        "contacts": "none",
        "analytics": "read"
      },
      "access_note": "Transactional email; no contacts or campaigns.",
      "sends": "yes",
      "sends_note": "through the API, as Mailchimp's own example shows",
      "safeguards": "What the server can do follows the API key: a key restricted to certain calls needs its AI Agents group enabled to use the MCP.",
      "watch_out": "Transactional only: Mailchimp documents no MCP server for its marketing platform's account data.",
      "summary": "Teams on Mailchimp Transactional (Mandrill) who want an agent to build templates, call the API and diagnose failed sends.",
      "capabilities": "Nine tools: build a template, list, describe and call any Transactional API endpoint, diagnose a failed send, check the account and get onboarding and integration help.",
      "checked": "2026-10-12",
      "source": "https://mailchimp.com/developer/transactional/guides/how-to-use-mailchimps-transactional-messaging-mcp",
      "more_sources": [],
      "publisher_row": false
    },
    {
      "id": "onesignal",
      "vendor": "OneSignal",
      "server": "official",
      "maker": "OneSignal",
      "status": "open beta",
      "connect": [
        "remote-oauth"
      ],
      "endpoints": [
        "https://api.onesignal.com/mcp/oauth"
      ],
      "access": {
        "segments": "change",
        "campaigns": "read",
        "templates": "change",
        "contacts": "change",
        "analytics": "read"
      },
      "access_note": "Reads past messages; sends through send_message.",
      "sends": "yes",
      "sends_note": "push, email or SMS",
      "safeguards": "send_message is flagged high-impact so clients can ask before running it; targeting and inputs are validated before a send, calls are rate-limited, and any client can be revoked from Connected apps.",
      "watch_out": "In open beta: an app may need access enabled before tools other than the utilities work.",
      "summary": "OneSignal apps that want an agent to send and analyze push, email and SMS, and to manage users, segments and templates.",
      "capabilities": "43 tools: send a push notification, email or SMS, manage users, subscriptions, segments and templates, export audiences, and check deliverability and email warm-up.",
      "checked": "2026-10-12",
      "source": "https://documentation.onesignal.com/docs/en/model-context-protocol",
      "more_sources": [],
      "publisher_row": false
    },
    {
      "id": "knock",
      "vendor": "Knock",
      "server": "official",
      "maker": "Knock",
      "status": null,
      "connect": [
        "remote-oauth",
        "remote-api-key"
      ],
      "endpoints": [
        "https://mcp.knock.app/mcp"
      ],
      "access": {
        "segments": "none",
        "campaigns": "change",
        "templates": "change",
        "contacts": "change",
        "analytics": "read"
      },
      "access_note": null,
      "sends": "yes",
      "sends_note": "by triggering a workflow",
      "safeguards": "At sign-in you choose which capabilities to enable, with data management and debugging off by default, and the server ships no delete tools, to prevent accidental deletions.",
      "watch_out": "A service token skips the consent screen and turns every capability on.",
      "summary": "Knock accounts that want an agent to build and test notification workflows, with capabilities picked at sign-in.",
      "capabilities": "Creates workflows and their email, SMS, push, chat and in-app steps, manages users, tenants and objects, triggers workflows, and can hand complex tasks to the Knock agent.",
      "checked": "2026-10-12",
      "source": "https://docs.knock.app/ai/mcp-server",
      "more_sources": [
        "https://docs.knock.app/developer-tools/agent-toolkit/tools-reference"
      ],
      "publisher_row": false
    },
    {
      "id": "intercom",
      "vendor": "Intercom",
      "server": "official",
      "maker": "Intercom",
      "status": null,
      "connect": [
        "remote-oauth",
        "remote-api-key"
      ],
      "endpoints": [
        "https://mcp.intercom.com/mcp",
        "https://mcp.eu.intercom.com/mcp"
      ],
      "access": {
        "segments": "none",
        "campaigns": "none",
        "templates": "none",
        "contacts": "read",
        "analytics": "none"
      },
      "access_note": "Also writes Help Center articles and internal notes.",
      "sends": "no",
      "sends_note": "notes stay internal",
      "safeguards": "The OAuth flow shows a consent screen listing what the client may do, and internal notes are visible to teammates only: no tool sends a reply a customer can see.",
      "watch_out": "Support data rather than marketing campaigns, and workspaces in Intercom's Australian region aren't supported yet.",
      "summary": "Intercom workspaces that want an agent to search conversations and contacts and keep Help Center articles current.",
      "capabilities": "14 tools: search and read conversations, contacts and companies, create and update Help Center articles, and add internal notes to conversations.",
      "checked": "2026-10-12",
      "source": "https://developers.intercom.com/docs/guides/mcp",
      "more_sources": [],
      "publisher_row": false
    },
    {
      "id": "posthog",
      "vendor": "PostHog",
      "server": "official",
      "maker": "PostHog",
      "status": null,
      "connect": [
        "remote-oauth",
        "remote-api-key"
      ],
      "endpoints": [
        "https://mcp.posthog.com/mcp"
      ],
      "access": {
        "segments": "change",
        "campaigns": "change",
        "templates": "change",
        "contacts": "change",
        "analytics": "change"
      },
      "access_note": "Cohorts stand in for segments; Workflows for campaigns.",
      "sends": "yes",
      "sends_note": "batch workflow runs, with a confirm token",
      "safeguards": "A header or query parameter limits a session to read-only tools, filters narrow it to chosen features, and a batch send needs the audience count and a confirm token the API checks, valid for 15 minutes.",
      "watch_out": "The full server covers all of PostHog, from SQL to deleting persons: filter it to the tools a task needs.",
      "summary": "PostHog projects that want one agent across analytics, cohorts, feature flags, experiments and messaging workflows.",
      "capabilities": "Tools across every PostHog product: queries and dashboards, cohorts and persons, feature flags and experiments, surveys, and Workflows that draft broadcasts, publish workflows and run batch sends.",
      "checked": "2026-10-12",
      "source": "https://posthog.com/docs/model-context-protocol",
      "more_sources": [
        "https://posthog.com/docs/model-context-protocol/faq",
        "https://posthog.com/docs/model-context-protocol/tools",
        "https://github.com/PostHog/posthog/blob/master/services/mcp/schema/tool-definitions.json"
      ],
      "publisher_row": false
    },
    {
      "id": "amplitude",
      "vendor": "Amplitude",
      "server": "official",
      "maker": "Amplitude",
      "status": null,
      "connect": [
        "remote-oauth"
      ],
      "endpoints": [
        "https://mcp.amplitude.com/mcp",
        "https://mcp.eu.amplitude.com/mcp"
      ],
      "access": {
        "segments": "change",
        "campaigns": "read",
        "templates": "none",
        "contacts": "read",
        "analytics": "change"
      },
      "access_note": "Cohorts stand in for segments; guides and surveys are read-only.",
      "sends": "no",
      "sends_note": null,
      "safeguards": "The server uses your Amplitude permissions, and admins can turn it off or scope it with read and write role actions, project by project, which apply even to org admins.",
      "watch_out": "MCP access is on for every user in an organization until an admin restricts it.",
      "summary": "Amplitude customers who want an agent to query and edit charts, dashboards, cohorts, experiments and the tracking plan.",
      "capabilities": "Reads and writes charts, dashboards, notebooks, experiments, cohorts, feature flags, metrics and the tracking plan's events and properties; guides and surveys are read-only.",
      "checked": "2026-10-12",
      "source": "https://amplitude.com/docs/amplitude-ai/amplitude-mcp",
      "more_sources": [],
      "publisher_row": false
    },
    {
      "id": "fromhello",
      "vendor": "fromHello",
      "server": "official",
      "maker": "fromHello",
      "status": "early access",
      "connect": [
        "remote-api-key"
      ],
      "endpoints": [
        "https://mcp.fromhello.io/v1"
      ],
      "access": {
        "segments": "change",
        "campaigns": "change",
        "templates": "change",
        "contacts": "change",
        "analytics": "read"
      },
      "access_note": "A journey your AI builds stays a draft until it is published; contact changes are notes and AI context.",
      "sends": "no",
      "sends_note": "no AI tool sends a message",
      "safeguards": "A journey your AI builds stays a draft until it is published, and publishing, pausing or deleting through an agent takes a confirmation step in your AI client.",
      "watch_out": "AI actions can change a live journey without a confirmation step: update_journey can set or clear a published journey's exit condition, which takes effect on the next event, and other edits, such as changes to a segment or a template, apply when saved. fromHello Cloud is in early access through the waitlist.",
      "summary": "Teams on fromHello that want their own AI client to build segments, journey drafts and templates and read results.",
      "capabilities": "59 tools, shared with the built-in assistant, for segments, journeys, templates, tracking and analytics; a journey is created as a draft, with every node, in one call. MCP calls are never metered.",
      "checked": "2026-10-12",
      "source": "https://fromhello.io/en#ai",
      "more_sources": [],
      "publisher_row": true
    }
  ]
}