I built LaunchBase to avoid ripping out auth, billing, and access control at month 3.
It’s a fork-once SaaS foundation focused on correctness over shortcuts: auth, Stripe billing, credits, roles, uploads - all wired explicitly.
Built for founders with infra scars.
@dirtytoken Yes RBAC + workspace inheritance is where auth stops being auth and becomes product logic.
I’m handling it with a workspace-first model and explicit permission mappings, not ad-hoc checks. Middleware alone doesn’t scale there.
Mind if I DM you to compare approaches?
@agusvrancic Sharp niche. Reducing blank screen friction is real, especially for founders who think faster than they type.
Is the hard part more the voice → structure step, or managing drafts / iterations once ideas exist?
@agusvrancic Makes sense early traction on X is always tricky.
A lot of builders I’ve seen focus on one tight audience first, sharing real value and shipping updates consistently, rather than trying to promote everywhere at once.
Curious what’s your target audience or niche for this tool?
@xdrewmiko@ysuckme Got it, makes total sense to pipeline the next one while waiting on formalities.
Is the MVP fully functional, or are you still ironing out infra like auth, roles, or background tasks?
@xdrewmiko@ysuckme Respect, that’s a real commitment.
Building two SaaS at once is intense. Curious what pushed you to run both in parallel instead of focusing on one first?
@its_juan_fern Nice - shipping two variants early is the right move.
Out of interest what you’re optimizing for between the two: clarity vs. urgency, or different ICPs entirely?
@Realtime0727 Shipping this fast in a 30-day MRR sprint is no joke.
Curious, what’s slowing you down more right now: product iteration, infra (auth/billing/jobs), or distribution?
In these sprints I’ve found infra friction shows up quietly until it suddenly blocks speed.
@kiam00944654 For a pre-launch SaaS, I’d test one thing:
Can a brand-new user get value without help?
That usually exposes onboarding gaps, auth friction, and hidden setup assumptions fast.
Curious what you’re most unsure about testing right now?
@killingclown678@meta_alchemist Yes,“Looks fine” is the most dangerous state.
If a system can’t surface when invariants are violated, you only find out after users or revenue are affected.
@killingclown678@meta_alchemist Exactly. Scaffolding is easy but correctness is not.
Most failures show up where tools assume safety instead of proving it.
@long_looong_ago Yup. Declines happen unclear state is what erodes trust.
When users don’t know what the system thinks is true, hesitation turns into churn.
@nandanalinson This is the class of bug that only shows up in real usage.
When backend logic and auth/access state drift even slightly, the product still “works” (just incorrectly) and that’s much harder to detect than hard failures.
@AsifAnsari27430 Yes.
That’s the tricky class of auth bugs, nothing “fails,” the timing is just wrong.
When state arrives out of order, UI issues are usually just the first visible symptom.
@gudluck_reuben That’s usually a good sign.
The nastier bugs tend to not show us as server errors but show up as a ‘valid’ state that’s subtly wrong and only shows up later in access or data leaks.
Good luck dude!