trip-planner

Generates deterministic trip itineraries, Excel workbooks, and HTML brochures from travel data.

2|1|Updated Jun 23, 2026
One-click install
npx skills add https://github.com/Kaidanov/grekai-skills-4all --skill trip-planner-kaidanov
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: trip-planner
Source: https://github.com/Kaidanov/grekai-skills-4all/tree/main/skills/trip-planner
Command: npx skills add https://github.com/Kaidanov/grekai-skills-4all --skill trip-planner-kaidanov

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires openpyxl, requests, mcp, and includes scripts (resource) and references (resource) and assets (resource) components.

What problem does it solve? Planning a multi-day trip means juggling flights, stays, attractions, routing, budgets, and bookings by hand. This Skill turns initial trip data into a complete day-by-day itinerary, a scored target bank of places, a 7-sheet Excel workbook with live budget formulas, and a printable HTML brochure — with the core planning done deterministically at zero LLM-token cost. ## Core Features & Use Cases - Deterministic itinerary engine: Derives bases from stays, clusters targets to the nearest base, allocates them across days by pace and opening days, routes each day with nearest-neighbour + 2-opt, schedules clock times, and runs conflict checks (car-after-flight, no-lodging, book-ahead). - Scored target bank with enrichment: Attaches Google Maps, Tripadvisor, and Google Reviews links to every place, plus stars, review counts, photos, and real drive-times via Google Places / Distance Matrix when a GOOGLE_MAPS_API_KEY is set (cached on disk; haversine fallback offline). - Three output formats: A 7-sheet Excel workbook (Itinerary, Attractions, Foodies, Budget with live formulas, Costs Remarks, Notes and tips, Prompt), a self-contained print-to-PDF HTML brochure, and an inline itinerary. Also ships as an MCP server and Claude Code hooks. - Use Case: Ask for an 8-day Tuscany family trip and receive a routed day-by-day plan, a budget workbook whose totals recompute when you edit any cell, and a brochure with star ratings and map links — plus automatic warnings like a car return scheduled too close to the homebound flight. ## Quick Start Ask the assistant to plan a trip, for example: plan an 8-day Tuscany family trip with flights, two hotel bases, and a budget workbook and printable brochure.

Frequently Asked Questions about trip-planner

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate a trip itinerary with this skill?

Provide flights, accommodation stays, and a preferences block (pace, interests, max drive time), optionally with candidate places. The engine clusters targets to bases, allocates them across days, routes each day with nearest-neighbour plus 2-opt, and outputs an itinerary, Excel workbook, and HTML brochure.

Does the trip planner work without a Google Maps API key?

Yes. Without a GOOGLE_MAPS_API_KEY it still builds Google Maps, Tripadvisor, and Reviews search links for every place and estimates drive-times via haversine at roughly 70 km/h. With a key it adds stars, review counts, photos, and real drive-times, cached on disk.

How many LLM tokens does trip planning use?

The core planning uses zero tokens — anchors, clustering, allocation, routing, scheduling, and conflict checks are pure algorithms plus cached API calls. The model runs at most twice per trip: parsing booking documents into events and writing short blurbs, both cached.

What conflict checks does the itinerary engine catch?

The rule table catches CAR_AFTER_FLIGHT (car return too close to the homebound flight's airport buffer), NO_LODGING (a night with no accommodation), and BOOK_AHEAD (a target tagged needs_booking that is not booked). New rules can be added as rows.

Can I use the trip planner as an MCP server?

Yes. The mcp/server.py file exposes the engine via FastMCP over stdio with tools trip_template, enrich_targets, distance_from, generate_plan, build_workbook, and build_brochure. Register it in Claude Desktop or .mcp.json with a python command pointing at the server file.

What Python dependencies does the trip planner require?

Only openpyxl is required to render the Excel workbook. The requests package is optional and needed solely for live Google Places and Distance Matrix calls, and mcp[cli] is needed only when running the MCP server. The engine itself is pure standard-library Python.