Agentic Website Builder 2.0 by Lokuma
Design, build, and run your site with a design agent harness
349 followers
Design, build, and run your site with a design agent harness
349 followers
Lokuma 2.0 is a design-aware agent harness for websites. Most AI builders can generate a first draft. But real sites need structure, taste, brand consistency, editing, publishing, forms, and ongoing updates. Lokuma connects planning, design, style, assets, site state, edits, and publishing into one agentic workflow — so your website feels designed, not just generated. Design, build, and run your site with agents.
Triforce Todos
Great one team, BTW, if someone eventually wants to hand this off to a developer or export it cleanly, what's the code quality like?
Design Agent by Lokuma
@abod_rehman Thanks! Yes, full export anytime.
It's agent-written code — modern stack, readable, componentized. Not hand-crafted by a senior engineer, but clean enough for a developer to pick up without a rewrite. The targeted-patching approach in v2.0 actually helps here: edits stay localized, so the codebase doesn't drift into spaghetti over many iterations.
Congrats on the launch, Mu! You said most AI website builders generate a great first draft, then leave when your AI ships v1 what's the first thing that breaks the second time you try to change it? for us it's always the hero section layout
Design Agent by Lokuma
@imogen_wallace Imogen, you nailed it. Hero is the canonical "regenerate-the-whole-section-on-every-edit" trap — change a headline, lose the image; nudge the CTA, layout falls apart. We pick a hero recipe early (cinematic / split / overlay-rail) so the agent edits inside it instead of over it. Won't claim it's perfect — hero is hard — but it should hold for the kind of edits that used to torch v2. Try it on one you've fought with, ping back?
Best,
Mu
Really thoughtful launch congrats! What's the most unexpected thing that broke during your own testing when editing an AI generated site a month later? That story would help us trust the runtime.
Design Agent by Lokuma
@owen_shaw2 Best story: our auto-repair watchdog.
We built it so when a site loaded with a blank #root after a successful build, the runtime auto-files a "fix this" chat. Sounded great in tests.
Then a user opened a project from weeks earlier. The iframe took a beat to fetch, watchdog saw empty root, fired auto-repair → backend treated it as a fresh build under that project ID → completely unrelated content appeared in the user's project slot.
Fix: a phase-edge tracker. Watchdog only arms after we've actually observed a build complete in the current session. Can't mistake "page loading" for "site broken" anymore.
Trust the runtime — but only when the runtime knows what state it's in.
Best,
Mu
Congrats, Mu! For SMBs who are not technical how do they tell Lokuma what their brand looks like? Do they upload a style guide, point to an existing site, or describe it conversationally?
Design Agent by Lokuma
@barnaby_lloyd Barnaby — all of those work today, plus a fourth. Choices for non-technical users:
Describe it in chat ("earthy, editorial, like a wine label")
Drop a reference URL in chat — agent extracts palette + typography without copying the layout
Upload a logo or screenshot directly in chat — vision picks up the look; logo persists as the brand anchor
Use the Style panel — side drawer with logo upload, brand text, palette / font swap, composition picker — for users who'd rather click than type
Most mix two: a sentence in chat + a logo upload (either route). Multi-page style-guide PDFs aren't auto-parsed yet (on roadmap).
Best,
Mu
Is the focus on UI only, or does it include UX too? From another comment and reply, if I ask it to "make it more professional" does it understand how the target audience will perceive "professional"?
Design Agent by Lokuma
@you_x_you_i Mostly UI, with UX scaffolding, such as section order, copy hierarchy,
conversion flow, form wiring and etc. What we don't do: actual audience
research.
"Professional" is routed by industry tag, not by what your audience
perceives as professional. For example, fintech loads certain design style + Space Grotesk × IBM
Plex Sans. Boutique hotel loads oxblood + gold + Cormorant.
Defensible defaults — not researched answers.
For SMB founders that's a fair trade-off — fast first cut they can react to, before they invest in the research/strategy layer (which is your work, not ours). Starting point, not the finish line.
Hope this helps : )
Best,
Mu
@mu_li Nice response, and I understand the justification. You're definitely building with the SMB audience in mind. I do like that you're using different UI libraries to cater to different industries. I wish you luck, and maybe in the future we can have a conversation to see where our products align.
Design Agent by Lokuma
@you_x_you_i Thanks and look forward to it!
Best,
Mu
The "operate" part is what catches my attention. Building a site is table stakes now — maintaining it, updating content, fixing broken stuff without touching code is where 90% of non-technical founders get stuck. Does the agent handle things like SEO meta tags and structured data automatically, or is that still manual?
Design Agent by Lokuma
@ytubviral meta + OG tags + sitemap + robots are handled automatically. Richer SEO (structured data, schema-driven enhancements) is exposed through chat — agent picks up what each page needs based on its content type. Beyond SEO: content edits, broken-link fixes, form repair all happen in the same conversation. The site stays editable, not a one-shot dump.
Best，
Mu
Honestly, the biggest headache with AI builders is trying to change one small line of text a week later and watching the entire layout drift into absolute chaos. If this actually keeps edits localized without turning the codebase into spaghetti over time, it’s a massive win.
Design Agent by Lokuma
@mithra_xavier Yeah, this is the headache we think about most. Some tech architecure for now:
1. Style edits (palette, font, anchor) bypass the LLM entirely — direct
token rewrite + rebuild in ~15s. So "change brand color a week later"
can't drift the layout, because the LLM doesn't run on that path.
2. Content edits are surgical old_string → new_string patches, and a
post-build audit flags drift (e.g. hero src pointing at a stale image
reference) so the agent can self-heal before you see it.
Not 100% solved. But worst-case revert is minutes, not weeks. Still
iterating on tighter guarantees.