Genuinely asking. Because I've had my fair share of frustrating failure moments with it. But after MUCH trial and error (and then some more), I’ve found some things that actually work for me:
- I’ve stopped skipping the planning phase.
Beginner's mistake. I thought stopping to plan every. single. idea. was slowing me down and boy was I wrong. Design Mode and Plan Mode exist for a reason and using them in the right order actually saves time.
Design Mode is great for fast prototyping, getting a feel for the UI before anything is wired up. Plan Mode is where you structure the project before the agent starts building.
The temptation of skipping straight to generating, is usually where things start going sideways. Taking a few extra minutes in Plan Mode means the agent has something solid to work from instead of improvising.
- The Verifier Agent is NOT optional.
We listen and we don’t judge, ok? I was ignoring it.
Turns out that when it asks for feedback, it's actually checking your code and looking for things you might have missed. I caught a few issues that I probably wouldn't have found on my own hehe 🫠.
- Lite Mode for small changes.
The little lightning bolt thingy. If you're making a minor adjustment, switch to Lite Mode first and it runs a lighter model. Faster and cheaper. No reason to use full capacity for a one-line fix.
Nothing groundbreaking really, but adding these to my workflow has changed my opinion and view on Replit. I’ve also put out a short video recently with some other little pro tips if you want to give it a look.
Any Replit habits that took you a while to figure out? Do you even use Replit?
Top comments (2)
The planning/verifier split is the habit that changes the whole experience. The biggest trap with coding agents is treating them like a faster autocomplete instead of a teammate that needs a clear job boundary and a way to check its own work.
What helped me most is separating the loop into: plan the change, make the smallest useful implementation, verify with the actual app/test path, then only broaden the scope if the result survives. When I skip that, the agent starts optimizing for “sounds complete” instead of “is actually shippable.”
Also agree on lightweight mode for tiny edits. Spending full-agent energy on a text tweak or one-line CSS fix is like starting a whole construction crew to move a chair.
heyy @alikhatersaibreakroom! love this. Feels like we're basically running the same playbook lol. Curious how you decide when to 'broaden the scope', is that a gut call or do you have a rule of thumb for it?