NewsletterLane speaks the Model Context Protocol. Point Claude, Cursor or any MCP client at one endpoint and your agent can manage lists, import subscribers, send campaigns and read engagement — all scoped to your organization by an API key.
What you can do with it
Not just a few API calls — the MCP server exposes the operations agents need to run real email work end to end.
“Draft this week’s digest from these links, send it to the Weekly list, and tell me how it did.” The agent builds, sends and reports — start to finish.
Hand the agent a spreadsheet of signups. It creates the list and imports everyone — deduped and suppression-aware, no manual mapping.
“How did last week’s campaign perform, and which list is growing fastest?” Plain-language questions become real queries against your lists and stats.
Keep lists in sync from your own app, CRM or a scheduled job — let an agent reconcile subscribers whenever something changes upstream.
The toolbelt
Five read tools work with any key. Three write tools need the Editor role or higher.
List your subscriber lists with their member counts.
List subscribers, newest first. Filter by list, search term or limit.
List your campaigns and their lifecycle status.
List sending domains with DKIM / SPF / DMARC verification status.
Get a campaign with engagement counts — delivered, open, click, bounce, complaint, unsubscribe — per unique recipient.
Create a new subscriber list.
Bulk-add up to 5,000 subscribers. Idempotent and suppression-aware; optionally add them to a list.
Dispatch a campaign to its list immediately. Runs on the durable queue + worker.
Connect
Create an API key in the app under Settings → API keys, then drop the config into your client.
Uses mcp-remote to bridge the remote server over stdio. Restart Claude after saving.
Every request carries Authorization: Bearer <key>. The key resolves to exactly one organization — agents never see another tenant's data.
Read tools work with any key. Write tools — create_list, import_subscribers, send_campaign — require an Editor key. Give read-only agents a read-only key.
Each tool is a thin wrapper over the versioned /api/v1 REST API. Validation, suppression and deliverability behave identically whether a human, your code, or an agent makes the call.
Grab an API key, paste the config, and your assistant is one prompt away from sending.