What is this?
This is a personal travel planning website for our family trip to Portugal in AugustβSeptember 2026. It covers 14 nights across three cities β Porto, Lisbon, and Aveiro β with day-by-day itineraries, hotel and restaurant recommendations, interactive maps, a planning checklist, and travel tips.
It was made for our family of four. It is not a commercial travel guide or affiliate site.
100% generated by Claude
Every line of code, every data file, every component, and every piece of content on this site was generated by Claude Code (Anthropic's AI coding assistant) through conversational prompting. No code was written by hand. The site was built iteratively over multiple prompts, with each prompt refining and expanding the previous result.
The technology stack β an Astro static site with TypeScript data files, scoped CSS, and client-side interactivity β was chosen and implemented entirely by Claude based on the requirements described in conversation.
How it was built: the prompts
Here is a summary of the prompts used to build this site from start to finish, to give you a sense of how generative AI can be used for travel planning through iterative conversation:
- Shift the dates
The initial itinerary had arrival in Porto on August 21. The first prompt changed the arrival to August 22, and Claude shifted every date across all three cities β Porto, Lisbon, and Aveiro β including individual day dates, transport dates, and the overview page.
- Add an anniversary restaurant
September 3 is our 20th wedding anniversary. Claude searched for fine dining restaurants in Aveiro suitable for the occasion, recommended Salpoente (Michelin-recognised, in a converted salt warehouse), and added it to both the itinerary and the restaurant list.
- Add PyCon Portugal
PyCon Portugal takes place September 3β4 in Aveiro. Claude added two conference days to the itinerary, updated the return-to-Porto schedule, added a departure day (Sep 5), and adjusted the total trip length from 13 to 14 nights across all pages.
- Add maps and links to all places
Claude added website URLs and Google Maps links to every activity across all three cities. This involved updating the Activity type to support optional
urlandmapUrlfields, updating the DayCard component to render clickable link buttons, and adding links to approximately 40+ activities. - Add overview maps
Each city page got an interactive map (using Leaflet and OpenStreetMap) showing all attractions, hotels, and restaurants as color-coded pins. Claude created a reusable MapSection component, added a MapPoint type, and populated coordinates for every point of interest across all three cities.
- Expand the travel tips
The original overview page had 6 brief tips. Claude expanded this into 4 categorized sections (Money & Payments, Weather & Packing, Getting Around, Practical Tips) with 19 detailed tips covering currency, tipping culture, what to pack, inter-city trains, public transit cards, rideshare apps, power adapters, language, and health/safety.
- Adjust the return schedule
The original plan had an evening return to Porto on Sep 4. A prompt changed this to a morning return on Sep 5, which meant updating the transport info, restructuring the last two days of the itinerary, and removing the need for an overnight in Porto.
- Add ice cream selfie spots
A personal tradition: taking ice cream selfies at tech conferences. Claude researched gelato shops in Aveiro and added suggestions to both PyCon days β Gelados de Portugal (with Portuguese-inspired flavours like ovos moles) and Gelataria Italiana.
- Build a planning checklist
Rather than building a separate Python backend (which was discussed and rejected as over-engineering), Claude built a checklist page within the existing Astro site. It has 8 categories and 48 items covering flights, accommodation, documents, PyCon registration, activity bookings, packing, and pre-departure tasks. Progress is tracked with checkboxes persisted in localStorage, with a progress bar and per-category completion counts.
- Portuguese color theme
The color palette was updated to reflect the Portuguese flag β green, red, and yellow accents throughout the site, from the navigation active states to rating stars, section labels, progress bars, and the hero gradient.
- Add this About page
This page itself was the final prompt β explaining what the site is and how it was made.
Tools used
- Claude Code (Claude Opus 4.6) β AI coding assistant by Anthropic, used for all code generation, web research, and content writing
- Astro β static site framework
- Leaflet + OpenStreetMap β interactive maps
- GitHub Pages β hosting