MCP Travel Planner Agent Team

A tool-using travel planner that combines MCP-connected services, live search, and structured itinerary generation into a single planning workflow.

MCP Tooling Workflow

Use real accommodation data, live distances, and search-grounded context to build an itinerary that feels operational, not aspirational

This project turns MCP from an infrastructure concept into a concrete product surface. The planner uses connected tools for Airbnb-style lodging data, Google Maps distance and travel-time calculations, and current web context, then packages the result into a detailed itinerary with costs, logistics, and calendar export.

Overview

MCP Travel Planner Agent Team is designed as a high-agency itinerary generator. Instead of returning generic travel advice, it uses live tools to assemble a more grounded trip plan: where to stay, how far places are from each other, what logistics look like in practice, and how the overall budget fits the user’s preferences.

What The Product Does

Collects destination, trip duration, budget, start date, and travel preferences from a structured Streamlit form
Uses MCP-connected tools to gather accommodation options and map-based distance information
Builds a detailed day-by-day itinerary with activity timing, transport estimates, and budget-aware recommendations
Adds practical travel context such as weather, local transportation, cultural norms, and packing tips
Exports the generated itinerary as an .ics calendar file for downstream trip execution

Implementation Details

Built as a Streamlit application with a single planning flow and sidebar API-key configuration
Uses Agno agent abstractions with an OpenAI GPT-4o planning model
Uses MultiMCPTools to connect to Airbnb-related MCP tooling and travel/location MCP services
Uses Google Maps data for distance calculation, travel-time estimation, and location-aware itinerary logic
Uses GoogleSearchTools for current weather, local research, restaurant context, and practical trip details
Includes a calendar-export helper that converts itinerary text into a downloadable ICS file

Why It Matters

The core product lesson here is that MCP becomes tangible when the user can feel the tools changing the quality of the output. Accommodation options become more concrete, distances become operational rather than guessed, and the itinerary reads more like something a real concierge would hand over than a purely synthetic travel summary.

Design Decisions

The planner is instructed to generate a full itinerary in one pass rather than turning into a long back-and-forth prompt session
MCP tools are used for the parts that most benefit from live external grounding: lodging and travel distances
Search fills in contextual details like reviews, weather, and travel tips without overloading the core planning logic
Calendar export makes the output immediately actionable beyond the chat-like interface

Role and Focus

Role: Solo builder focused on tool-using agent UX, itinerary quality, and MCP-powered product workflows.

Tech Stack: Streamlit, Agno, OpenAI GPT-4o, MultiMCPTools, Google Maps integration, Google search, iCalendar export.

Category: MCP agents, travel planning, tool orchestration, real-time data workflows.

Positioning: Think "travel concierge with live tools," not just a conversational travel recommender.

Thumbnail Alt Text

MCP-based travel planning interface showing destination inputs, budget-aware itinerary generation, live location reasoning, and downloadable calendar output.