Not long ago, I decided to commandeer my data that I have entered into other media tracking applications like Letterboxd and Goodreads. Nothing wrong with either of these applications but I wanted a record of each piece of media I consumed and all the details that go with it.

While these applications probably are not going anywhere, if they stopped providing their service for whatever reason, my data would also be gone. Decided to copy my data from these applications over to my PDS. I currently do not host my own PDS, choosing to use my default option: social.bsky.unceman

My book data was fairly easy. I exported by Goodreads data into a CSV, then imported into Bookhive.buzz, which is ATProto aware. Although the imported book data was not in social.bsky.unceman it was still in my PDS by way of buzz.bookhive.book .

The movie data was a different story. Although I did a similar CSV export/import, it required going back to enhance the movie attribute data because Letterboxd essentially archives older entries from their exports with limited details. I tried to create something in Pipedream to pull the data from Letterboxd, then create a custom record in my PDS. I failed over and over doing it this way. Pipedream proved to be unforgiving to my remedial coding skills.

Finally, I leveraged Claude Code and it was spectacular in its troubleshooting persistence which offset the errors that it generated in its provided code. Turned into a real back and forth conversation between Claude and me, where we both had to recognize what was and was not working. A very basic explanation is that I established the attributes from my personal Letterboxd RSS feed, then reused those attributes. The final step was to populate missing attributes from TMDB. Ended up with all my data in a custom location called social.bsky.unceman.film

After we were finished, Claude was nice enough to ask if I wanted a front end to my media data. He generated an index.html file and I put it on Netlify to host my single HTML file.
https://unceman.netlify.app/

All that is missing is my music scrobble from Spotify/Last.fm. I have a Pipedream trigger that runs monthly to create a Bluesky post with my Top 5 most listened to songs on Spotify. My next project will be to incorporate the music data a little better, but that will take some figuring out.