Activity
Mon
Wed
Fri
Sun
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
What is this?
Less
More

Memberships

Vibe Coding School

277 members • $30/m

17 contributions to Vibe Coding School
My Stack - Thoughts?
āœ…Ā Claude CodeĀ - AI that builds from prompts āœ…Ā ReplitĀ - Instant cloud IDE with deployment āœ…Ā GitHubĀ - Version control and professional workflow āœ…Ā RailwayĀ - Full-stack app hosting āœ…Ā SupabaseĀ - Database + auth + real-time + storage I will change to Lovable for my next project on Zacs reccomendation
2 likes • 17d
That's a solid stack. Have you tried out cursor instead of claude code? Then again, if it gets the job done that's the most important thing.
šŸ’ø Introducing Our Affiliate Program — Earn 50% every month for every member you invite into the community
Lovable Vibe Coding is now $30/month to power bigger challenges + prizes. How you earn (recurring): Go to the community page → click ā€œInvite peopleā€ → copy your link. If they join via your link, you earn 50% of the communities monthly fee every month per each member that joins the community through your link: - When we hit 280 members, price becomes $40/month You must use your invite link to get paid. New in the Classroom - Free: Lovable Zero to Ship in 5 Days: https://cedar-salute-3ea.notion.site/Lovable-Zero-to-Ship-in-5-Days-25a47b7b39aa80baab12d1f4849cf973 - New playbooks in classroom: 1. Upwork: Land Your First Client 2. Ideation → PRD → Lovable + Prompting Mastery 3. Secure Supabase with Lovable 4. SEO in Lovable šŸ”„ New challenge dropping very soon. Share your invite link on IG/TikTok/X and DM 3 friends to start earning every month. šŸš€
3 likes • Aug 26
This is awesome!
How to Get Free Traffic from ChatGPT: The Complete Playbook
Found an excellent post on AI coding lessons - here's my summary with a few additions from my own 8 months building with Lovable AI traffic has grown 1,367% since January 2024. ChatGPT now sends more referral traffic than Reddit and LinkedIn combined. Here's how to capture your share. 1. Technical Setup (Do This First) Unblock AI Crawlers: Disable AI blocking in Cloudflare, Akamai, Fastly, Imperva, Wordfence, DataDome Whitelist: GPTBot, PerplexityBot, ClaudeBot Kill JavaScript challenges for verified bots Serve critical content in server-side HTML (GPTBot and PerplexityBot ignore JavaScript) Robots.txt Permission: text User-agent: GPTBot Allow: / User-agent: PerplexityBot Allow: / User-agent: ClaudeBot Allow: / Essential Files: Add JSON-LD schema to articles/FAQs (Copilot reads it verbatim) Include visible "Updated YYYY-MM-DD" timestamps Create /llms.txt with your top 50-100 URLs + one-line summaries Ping IndexNow after every publish - Bing/Copilot will re-crawl within minutes 2. Content That Gets Cited Answer-First Structure: 40-70 word answer box under your H1 (prime feature-snippet fodder) Expert quotes in each section (+41% citation lift) Fresh dated stats (+30% citation boost) Keep paragraphs ≤300 tokens, break with H2 every ~250 words Use clear, standalone subheadings High-Performance Content Types: Comparison guides ("what's the best...") Step-by-step tutorials with numbered processes FAQ sections with schema markup Data-driven insights with specific statistics Content Strategy Details: Go easy on hard-sell CTAs in paragraphs you want quoted Seed Reddit threads AND Stack Overflow discussions Keep each fact/stat within single vector chunks 3. Platform-Specific Strategies ChatGPT: Rank high on Bing, use IndexNow for fast re-crawling, neutral Wikipedia-style tone Perplexity: Post authoritative Reddit responses (47% of answers cite Reddit), avoid JavaScript-heavy pages, pure HTML only Google AI: Traditional SEO + structured data, quarterly freshness updates
2 likes • Aug 23
This is an interesting read, thanks Zac.
LoRA training for images?
Has anyone done LoRA training for images? I'm creating a Pikzels thumbnail clone and Lovable is getting stuck with the LoRA training (not lasting long enough to get my likeness). Has anyone done LoRA training manually? I'm using Replicate.
2 likes • Aug 12
Based on the documentation, Lovable's Replicate integration is designed for inference (using existing models) rather than training custom LoRA models. It seems as if the platform focuses on calling pre-trained models like Flux Schnell for image generation using prompts and API calls. For your thumbnail clone, you'll need to handle LoRA training outside of Lovable/Replicate and then potentially deploy the trained model back as a custom Cog model. The documentation suggests using Replicate's Playground to test different existing models and refine prompts before integration.
How useful is the chat?
Hello everyone. I was just thinking about how useful the chat function actually is. I understand that it makes sense to enter the various tasks you want to implement in the chat first to check that Lovable has understood the requirements correctly. In my case, I have to say, I rarely if ever make changes, but press Implement afterwards. The question I ask myself is, am I not unnecessarily incurring double costs because lovable carries out planning beforehand anyway which is just not displayed to me if I click on Implement immediately? šŸ¤“
3 likes • Aug 12
You're not paying double costs. Chat Mode charges 1 credit per message for planning/discussion, while Default Mode charges 0.50-2.00+ credits based on complexity when it immediately implements code changes. If you rarely need to refine the AI's understanding and usually implement right away, Default Mode is likely more cost-effective for your workflow since you skip the planning credit and go straight to implementation.
1-10 of 17
Andrea Nicchia
3
39points to level up
@andrea-nicchia-7220
learning one step at a time

Active 17d ago
Joined May 16, 2025
Powered by