Claude Code / Cursor / MCP — Make calls: how to connect

Which AI do you use

What should it do

Look things up Run campaigns Publish work Make calls

Claude Code / Cursor / MCP · Lingsheng voice agent Account needed

Connect to the Lingsheng MCP with an access token to configure AI reception; place calls, read records and receive webhooks via the open API

Step 1 of 3

Step 2 of 3

Claude Code: run this, replacing the token with yours

Config / command

claude mcp add --transport http lingsheng https://aicall.lingshengai.com/api/mcp \
  --header "Authorization: Bearer lsk_your_token"

Step 3 of 3

In Cursor, add this to ~/.cursor/mcp.json (or the project's .cursor/mcp.json), with the same Authorization header Written from the vendor docs; not yet tested on a real client

Config / command

{
  "mcpServers": {
    "lingsheng": {
      "url": "https://aicall.lingshengai.com/api/mcp",
      "headers": {
        "Authorization": "Bearer lsk_your_token"
      }
    }
  }
}

Once connected

AI AI Assistant

Look at this store's AI reception and make the greeting shorter

✓ Used Lingsheng voice agent

Done — the new greeting is live:

Before"Hello, thanks for calling. We are open from 10 a.m. to 10 p.m. How can I help you today?"
Now"Hi, front desk here — booking a table or asking about a promotion?"

The next call uses it. Want to hear it first? I can call you.

Message…↑

Illustration

Say“Look at this store's AI reception and make the greeting shorter”

Full developer docs

← Everything your AI can do with us