Google Travel Planner MCP icon Travel Planning MCP

Google Travel Planner MCP

Use one skills-based MCP call to coordinate flights, hotel discovery/pricing, and map-aware trip planning in a single assistant workflow.

  • check_circle One public tool to run a full travel-planning agent flow
  • check_circle Coordinates Google Flights, hotel pricing, and map context together
  • check_circle Works in ChatGPT, Codex, and other MCP-capable clients
Google Travel Planner recommendations demo

FAQ

Common questions teams ask before integrating Google Travel Planner MCP.

How do I install this?

Set up your Infoseek account to get your app URL, then paste that URL into your preferred AI client.

What does this MCP do?

It runs a travel-planning agent that coordinates flight search, hotel planning, and map context inside one flow.

How many tools are exposed?

One public tool: run_agent. You provide a travel prompt and optional file URLs.

Can I use this for multi-part trip requests?

Yes. It is designed for multi-step prompts covering flights, hotels, and location tradeoffs in one run.

Can I pass files at runtime?

Yes. Use the optional file_urls array if your workflow needs runtime documents or references.

Tools

This MCP intentionally exposes a single agent runner interface to keep integration simple.

Tool

run_agent

travel_explore

Runs the skills-based travel planner using its bundled skill configuration, pinned MCP apps, and optional runtime files.

Required input

  • prompt

Optional input

  • file_urls[]

Output

  • content[]
  • isError

Response JSON

A typical tool call returns agent output as MCP content records.

{
  "content": [
    {
      "type": "text",
      "text": "Flights shortlist, hotel recommendations, and map-focused trip guidance..."
    }
  ],
  "isError": false
}

Quickstart + Code

One-call workflow

  1. 1. Pass a detailed travel prompt to run_agent.
  2. 2. Optionally attach file_urls for trip constraints or references.
  3. 3. Render the returned guidance as your trip-plan response.
const result = await mcp.callTool("run_agent", {
  prompt: "Plan a 4-day NYC trip with nonstop flights and hotels in SoHo",
  file_urls: []
});

return result.content?.[0]?.text || "";