Shipwright Dashboard banner

Shipwright Dashboard

3 devlogs
5h 36m 43s

This is the dashboard shipwrights use to review your flavortown projects :)

This project uses AI

Used gemini for debugging and some code implementation suggestions

Demo Repository

Loading README...

Obay

The final devlog of Flavortown…

Attachment
11

Comments

Simon
Simon 30 days ago

yooooooooo!

sage
sage 30 days ago

LAST DEVLOG OF FLAVORTOWN YOOO

PianoMan0
PianoMan0 30 days ago

o7

Obay
Obay 30 days ago

Salute from the Shipwrights

Shuflentity
Shuflentity 30 days ago

🐸🚀

Ghost of Nova
Ghost of Nova 30 days ago

bye flavortown :(

RespectableDot
RespectableDot 30 days ago

thank you for everything flavortown!

Anirudh
Anirudh 30 days ago

o7

Kuzu
Kuzu 29 days ago

BE A SHIPWRIGHT DIE A SHIPWRIGHT

Elliott-entity
Elliott-entity 29 days ago

o7 shipwrongs for life

Obay

I’ve implemented the theme logic to switch themes and set up CSS variables so they update automatically based on the selected theme.

I started by adding my favorite theme, which I like to call paper It’s a light-mode theme, but don’t worry, I’ll add a midnight theme and some dark themes later on (dark users, I see you 👀)

There’s still a lot left to theme across other pages, and Eric’s code is packed with many different colors that I’ll need to convert into custom variables. Wish me luck.

Here’s the commit for this devlog :)
see the amount of colors/pages changed sob-ios

0
Obay

So basically, the shipwrights need to check some things like the README and the demo URL and make sure they are working. They can miss some things, and there are some projects that aren’t really easy to understand how they’re working if they’re not documented well.

So this AI first takes:

  • project_name
  • project_type
  • readme_content
  • demo_url
  • repo_url

Then it passes all of these to Gemini 3 Flash with a very large prompt. You can take a look at it here:
https://github.com/hackclub/shipwrights/blob/main/sw-ai/Source/helpers.py#L129

This is the part of the sw-ai that is written in Python.

For the dashboard part, it will get all these variables from the database, but it will get the README content from the raw URL, which is easy to fetch. This is handled by a server action since we use Next.js in the platform.

And finally, it adds the newly generated summary to the database and shows the AI summary in the UI.

Attachment
0