Back
LDMLDM Academy
MCP flow

The AI-driven SDR process: MCP flow

The platform has two modes. This guide covers the first one: you bring your lists, LDM handles delivery, and the whole process is driven by your AI agent or a chat with an MCP connector. From sign-up to replies in the Inbox.

LDM Academy·2026·~8 min read

Chat or agent

Work the way you like

There was a wave when everyone lived in chats. Now people are moving to agents. For us it's the same thing: the MCP connector plugs into a chat (Claude Desktop, Perplexity) and into an autonomous agent (Claude Code, Cursor, any MCP client) alike. A person in a chat gives tasks in plain words; an agent does the same on its own.

Either way, the same LDM sits on the other side: the same tools, the same expert layer, the same legal guardrails. The interface is a matter of habit, not capability.

Chat or agent doesn't matter. What matters is that the process follows a proven flow, not gut feeling.
Our edge

The LDM MCP agent doesn't just execute — it accompanies

However smart your agent is, it doesn't see the whole context: deliverability state, domain history, the typical pitfalls of each stage. So LDM answers not as a bare API but as a market expert: every response carries an _expert block — what to do next, the typical flow, where people usually fail.

Think of it as a consultation: a lawyer and a doctor can each be brilliant alone, but exchanging opinions in the moment prevents expensive mistakes. Your agent makes the decisions — our layer keeps reminding it what's going on around. The capability map and end-to-end scenario live at /api/v1/agent-guide; the agent reads it itself.

A person can do all of this alone. Our job is to keep them — and their agent — from failing where everyone fails.
The launch flow

From zero to replies — eight steps

This is the recommended path — one of several, but the one that genuinely makes life easier. Nothing is invented here: every step has an MCP tool and a documentation page behind it.

  1. 01 · Sign-up
    Automatic, from a chat or by the agent. Terms acceptance right away: the legal layer starts at step one.
  2. 02 · Sending accounts
    Add the mailbox — or pool of mailboxes — the outreach will be sent from. The platform checks the connection and DNS and tells you what to fix.
  3. 03 · List
    Upload your list (CSV, API, or straight through MCP). De-duplication, stop lists, and suppression run automatically.
  4. 04 · List confirmation
    Confirm the source and the lawful basis for contact. Not bureaucracy — it's what separates outreach from spam.
  5. 05 · Enrichment
    The SDR layer suggests which variables to enrich the list with for your segment — so the email is about the recipient, not about you.
  6. 06 · Creative
    Preparation, sandbox on your own contacts, a test send to yourself — then approval. Nothing goes out without an explicit sign-off.
  7. 07 · Sending
    The campaign runs with continuous inbox-placement checks on a seed pool. Inbox is billed; Spam and non-delivery are not.
  8. 08 · Dialogs
    Replies land in a single inbox: classification, answers, interested contacts become leads. The agent runs the dialogs together with you.
Every stage has a legal layer. You don't have to remember it: the platform won't let you skip it.
Getting started

You can start right away

  • 01Sign-up is automatic — the agent completes it for you from the chat.
  • 02You start with $50 on the balance: enough for the first campaigns and checks.
  • 03A free seed pool for inbox-placement checks (with seed-count limits on the free tier).
  • 04Honest billing: you pay only for emails that actually reached the Inbox.
An MCP agent can walk the whole path — from sign-up to the first campaign — in one session. You confirm the launch.
Connect

Connect right now

Copy the config for your client — or just hand the agent a ready-made prompt and it does the rest.

Per-client configs
  • Claude Desktop — claude_desktop_config.json — the snippet below. Restart the app and the LDM tools appear on their own.
  • Claude Code — One command: claude mcp add ldm -e LDM_API_KEY=ldm_… -- npx -y --package=@live-direct-marketing/sdk@latest ldm-mcp
  • Perplexity (Mac) — Settings → Connectors → Add Connector → Simple tab: the same npx command with the key in env.
  • Cursor and other MCP clients — The same snippet in the client's MCP servers config.
{
  "mcpServers": {
    "ldm": {
      "command": "npx",
      "args": ["-y", "--package=@live-direct-marketing/sdk@latest", "ldm-mcp"],
      "env": { "LDM_API_KEY": "ldm_..." }
    }
  }
}

Prompt for your agent — copy and paste

Connect to LDM over MCP (use the config from the guide; ask me for the key — or register a new account). Read /api/v1/agent-guide. Then walk me through the first launch: add a sending account, import my list, suggest variable enrichment, prepare a creative, send a test to my address, and show campaign readiness. Launch only after my confirmation.
Web chats: ChatGPT, claude.ai, Gemini (hosted remote server)
https://api.live-direct-marketing.online/mcp

These clients can’t spawn a local process — connect them to our hosted server at https://api.live-direct-marketing.online/mcp. In claude.ai / Claude Desktop add it as a custom connector with a request header Authorization: Bearer ldm_.... In ChatGPT (developer mode), where custom headers aren’t supported, use the personal URL https://api.live-direct-marketing.online/mcp/k/<your-key> with authentication set to None — and treat that URL as a secret.

Documentation

  • Developer portal — developers.live-direct-marketing.online — API, keys, MCP, A2A.
  • Platform help — live-direct-marketing.online/help — the whole interface, section by section.
  • Agent map — /api/v1/agent-guide — the end-to-end scenario, read by agents automatically.

Questions? Write to us — we'll reply personally

No funnels, no forms. We'll tell you how this flow maps onto your case.

From Artyom, the developer: I value networking and plain human conversation — voice calls, messengers, however works. No LDM pitch, no agenda. Just glad to connect.