How This Developer Built Cursor Directory in 3 Hours to $35K/Month
TL;DR: Pontis and Victor were bootstrapping main startup Midday when Pontis discovered gap on flight: no central place to find Cursor rules (just scattered GitHub gists/forums). Called Victor immediately upon landing in France. Victor spun up Figma design while Pontis created Next.js project with hard-coded JSON file of rules found online. Had working prototype in 30 minutes on Vercel, bought cursor.directory domain, completed site in 3 hours total. Launched on X (1M impressions on initial post), hit HackerNews front page, YouTubers started covering it organically. Now: 49K registered users, 2.2M unique visitors since launch, $35K/month revenue, 99.8% profit margins, $525/mo operating costs, 3 hours/month maintenance. Added MCP support months later creating second adoption wave. Revenue model: paid job ads, featured MCP listings, cursor rules generator tool. Tech: Next.js, TypeScript, Shadcn components, Resend, Vercel hosting, Polar payments. Success factors: developer following from building in public on X, tech-heavy design approach, perfect timing with Cursor's growing popularity, reused design system/components from previous startup.
Key Insights
- 3-hour build constraint combined with immediate launch prevented overthinking
- Directories still work in 2025 if you catch right audience at perfect timing
- Design differentiates in AI-generated world where most sites look identical
- Reusing design systems from previous projects dramatically accelerates new builds
- Building in public creates developer following that amplifies organic launch
Actionable Takeaways
- Spot gaps during everyday activities - Pontis found idea during flight research
- Launch immediately when you spot opportunity - called co-founder same day
- Reuse components and design systems from previous projects for speed
- Post on X, HackerNews, and let organic coverage from YouTubers amplify
- Add new features (like MCP support) to create second waves of adoption