February 18, 2026
Therapy for your app idea
Stop prompting. Let the AI interview you to build specs
“Let the AI grill you” has non‑devs cheering, devs side‑eyeing
TLDR: A tool lets AI interview you to create a full product spec before coding. Non‑tech founders love the clarity, while developers say they already do this and raise privacy worries; links to similar tools suggest the idea is useful but not entirely new.
A new tool promises to fix the prompt chaos by flipping the script: don’t prompt — let the AI interview you and spit out a full product spec. Hacker News turned into confession hour fast: enha admitted, “I’m not a developer,” and described the dreaded “bug‑fixing loop” after one‑sentence prompts. Cue the relief: non‑tech founders love the idea of an AI “project manager” asking the right questions before any code is written.
But the crowd wasn’t all hugs and high‑fives. Privacy alarm bells rang loud: xnorswap pointed out there’s no privacy policy or about page, and worried the tool might “run away with my idea.” Meanwhile, seasoned builders flexed: d--b already asks AI to “list the questions you need to build X,” because large language models (LLMs) make assumptions when you don’t spell it out. Others dropped receipts: ravroid uses a custom GPT to interview themselves, then turns the spec into a plan for Codex; ochronus linked similar tools like GSD and spec‑kit.
The sample output — a chunky “LexiQuest” spec with goals, flows, and data models — had folks calling it requirements therapy. The vibe: founders are thrilled to get clarity, devs are skeptical about trust and novelty, and everyone’s joking that the AI just became your boss.
Key Points
- •The tool replaces manual prompting with an AI-led, adaptive interview to gather requirements.
- •It generates a complete engineering specification ready for development or execution in Codex/Cursor.
- •Output covers goals, user flows, data models, key screens, boundary conditions, error handling, and UI/UX.
- •A sample PRD (LexiQuest) demonstrates features: intelligent input, AI analysis engine, test-driven review, and gamification.
- •The sample includes technical architecture (service components) and a detailed data schema using UUID, JSONB, Enum, and Timestamp.