another personal project: Compra Clara
i’m building it to compare grocery costs across shops, including pack sizes, offers and travel costs
Python + SQLite
turns out “where is my shopping cheapest?” has a few more conditions than i expected
¡Hola, Region of Madrid! 🇪🇸 We’re thrilled to announce that we’re coming to Spain, beginning with Madrid. This will be our first region in Spain and second in the European Union. We plan to launch fully autonomous ride-hailing in Madrid in 2027.
Sign up for future updates: https://t.co/gOp8VmhqCb
¡Hola, Comunidad de Madrid! 🇪🇸 Nos emociona anunciar que llegamos a España, empezando por Madrid. Esta será nuestra primera región en España y la segunda en la Unión Europea. Planeamos lanzar nuestro servicio de ride-hailing totalmente autónomo en la Comunidad en 2027.
@askOkara learning from edits is the interesting bit here. can users mark an edit as “just for this post” so a one-off change doesn’t become a permanent style rule?
@Henryf1w yeah, explaining an error message is a good example. AI can give you the context straight away. checking the fix against the docs for your exact version is still worth doing
@tornikegomareli making the destination feel like a place instead of a device name is a nice touch. “put this file over there” is a pretty natural way to think about sharing
@jsnnsa@OneHitKillDev the delay after the fifth block is such a good example. the code can match the brief and still feel off. actually playing what you build catches things a checklist won’t
@filicroval that price difference makes experimenting a lot more practical. i’d love to see the full run cost, retries included. that’s the number i’d use to decide whether to build with it.
@doooyle the custom screensaver is the bit i like most. calendar and weather on a screen you already own is a useful little project. doesn’t need to become a whole startup.
@thsottiaux good to see the refunds included in the recovery. a clear way to check whether your account was affected would help too. people need to know what happened to their own usage.
@devteamdrew@AnthropicAI@claudeai makes sense that the dev tools need to catch up too. generating more code puts more pressure on builds, merges and debugging. those are engineering problems worth talking about.
@AnupamHaldkar what triggers escalation in your setup: a failed check, a separate evaluator, or the model’s own confidence? that seems like the tricky part to get right.
the OpenCode recovery is a reminder: spending controls belong in the product too.
if an agent can spend API credits, i want a budget, usage i can trace, and an easy way to stop it.
useful even when nothing goes wrong.
Our team worked with OpenCode to have their users who are also OpenAI API customers back in business and any credits refunded which would have been drawn in the attack.
Also, you can request Daybreak Blue access to use our cyber permissive setup for defense. Do not wait until you have to deal with an incident. We’re here to help.
Gym Log fills in the previous weight and reps before you log a set.
small detail, but it means less typing between exercises.
you can change the numbers when needed. otherwise, one tap and back to the workout.
@minchoi does your plan include the exact files to change and how to check the result? that’s the detail i’d want before handing it over to the coding agent.
@hqmank this is where a clear message inside the app would help. “your plan doesn’t include this” and “update the app to use it” need very different fixes. a missing model in the dropdown explains neither.
@felix_trz making it something people can move around in is a nice way to explain the scene. you can look at the bit you’re curious about instead of replaying the same camera angle.
@neil_xbt “use the existing auth flow” is a good example. obvious to the person writing the brief, not so obvious if there are three old implementations in the repo. naming the file would save a lot of guessing.
@dannypostma turning the conversation into a task makes sense. i’d keep a small “done means…” section in each one, so the useful details from the brainstorm don’t disappear when the coding starts.