Driving a whole voice-AI platform from a single prompt

·1 min read ·#MCP#OAuth#AI agents#TypeScript

I just shipped the complete MCP integration layer for OmniDimension: two services, 49 tools, a full OAuth 2.1 server, published to npm and live in production.

Any MCP-compatible AI client can now operate an entire OmniDimension account in plain language:

“Create a voice agent that transfers to a human when asked, attach this PDF as its knowledge base, give it a phone number, and start a bulk campaign to my contact list.”

That single prompt drives agent creation, knowledge bases, phone provisioning, and bulk calling, automatically.

Two services, two entry points#

I built it as two services because developers and end users need different doors in.

  • A local npm package (@omnidim-ai/mcp-server) for developers who want it directly inside their IDE.
  • A hosted OAuth service (omnidim-mcp-cloud) for users who just click “Connect” in Claude or any other MCP connector.

The thing I cared about most: the MCP tools stay synced with our public APIs automatically, so new platform capabilities show up to AI clients the moment they ship. No second copy of the API surface to keep in step by hand.

Developer experience#

A tool layer is only as good as the day you have to debug it, so I spent real time there:

  • proper diagnostics tooling
  • local debug logs
  • privacy-safe telemetry
  • production CI and testing before every release

The pace tells the story: 49 tools, 2 services, OAuth 2.1, 12 npm releases in 7 days.

Try it#

It works in any MCP client. The docs walk through connecting Claude Desktop and controlling real agents:

If you try it, leave a star or open an issue. We’re iterating fast.

Originally posted on LinkedIn, with a silent demo of Claude driving real agents.