Pinbucket

A dead-simple way to save places you find on Instagram Reels. Instead of screenshotting and forgetting, you type the name, Wikipedia fills in the rest, and it lives on a Pinterest-style board.

Why this exists

Instagram is full of beautiful places — fjords in Norway, mountains in Kazakhstan, villages in Italy — but saving them is a mess. Screenshots, bookmarks, DMs to yourself. Pinbucket gives you one place to collect them. Type a name, get the image and description from Wikipedia, add a note, and you're done.

No accounts. No API keys. No scraping Instagram. Just Wikipedia's free REST API doing the heavy lifting.

What you get

• Live map hero — a full-width OpenStreetMap view (Leaflet) sits at the top of the page and fades into the board. Every saved place with coordinates gets a subtle pin; hover for the name, click to open the card. Tiles switch between light and dark with the theme.
• Wikipedia-powered cards — type a place name, pick from 5 search results, and Wikipedia auto-fills the image, description, and coordinates. Multiple images are pulled from the Wikipedia media gallery so you can flip through them in the modal.
• Pinterest-style board — full-image cards with a gradient overlay and place name. Hover to reveal the description and your notes. Pure CSS masonry grid, responsive down to single-column on phones.
• Tags & references — tag places with things like "hiking", "norway", "waterfall". Save the Instagram Reel URL (or TikTok, YouTube, whatever) as a reference so you can find it later.
• Dark mode — toggle in the header, sticks between sessions.
• Manual fallback — if Wikipedia doesn't have your place, fill in the name, image URL, and description yourself.
• Everything is editable — click any card to open the full view, then hit Edit to change tags, references, or your note anytime.

Running it

pip install flask requests
python app.py

Opens on localhost:9875.

Running the tests

pip install pytest
python -m pytest tests/ -v

23 tests covering search, preview, CRUD, tags, references, partial updates, edge cases. No network calls — all Wikipedia requests are mocked.

Stack

Flask + SQLite + vanilla HTML/CSS/JS. No frameworks, no build step, no npm. Just a couple of Python dependencies and a handful of static files.

API

Endpoint	What it does
GET /api/search?q=...	Wikipedia search, returns 5 candidates
GET /api/preview?title=...	Wikipedia summary + media gallery for a title
GET /api/locations	All saved cards
POST /api/locations	Save a card (with tags, refs, image_urls)
PUT /api/locations/<id>	Edit tags, references, note, or other fields
DELETE /api/locations/<id>	Remove a card