How was outbid.lol built? Grok 4.6, Cursor, and the tech stack

Which AI model and editor did the founder name?

Jonathan Wilke’s direct August 24, 2026 X post says outbid.lol was “100% built with Grok 4.6 in Cursor.” This is the strongest primary evidence for Grok 4.6 as the AI coding model and Cursor as the development environment.

“100% built with AI” does not prove zero human involvement. The founder may have directed product decisions, prompts, code review, testing, deployment, and incident response. The post is not an independent audit of generated-code percentage or a complete development log.

Was outbid.lol really built in three hours?

Relve, Superframeworks, and other independent launch analyses describe a roughly three-hour build. The official About page confirms the August 19, 2026 launch at 11:08 PM, but it does not publish a development start time or stopwatch record.

The defensible wording is therefore “reported as a roughly three-hour build.” Visible build time also excludes the founder’s prior product experience, reusable components, accumulated audience, payment-integration knowledge, and maintenance after launch.

• The launch timestamp is available from a primary source.
• The roughly three-hour figure is repeated by independent sources.
• Prompt and commit timelines are not public.
• Post-launch maintenance, moderation, and infrastructure work may fall outside the three-hour claim.

How much of the technology stack is verified?

The founder’s DDoS statement supports Vercel use, the current Terms name Dodo Payments for checkout, and the About page documents a move from Vemetric analytics to DataFast. Older launch material naming Polar shows that a payment component changed after launch.

The reviewed official pages do not identify the application framework, database, ORM, queue system, cache layer, or source repository. Third-party guesses about Next.js, Postgres, or a specific boilerplate should not be published as a confirmed tech stack without primary evidence.

What does the case prove about AI product development?

The case shows that an experienced developer can use AI-assisted tools to ship a tightly scoped distribution idea quickly. Choosing Grok or Cursor alone does not guarantee virality, payment safety, regulatory compliance, scaling resilience, or revenue.

The reusable lesson is to keep scope small without skipping signed webhooks, idempotent payment fulfillment, moderation, rate limits, observability, and original positioning. Generating a screen quickly and operating a reliable commercial service are different tasks.

• AI can compress implementation time.
• Distribution and an existing audience remain separate advantages.
• Architecture guesses are not verified without published code or primary disclosure.
• Speed does not remove security and operational controls.