Virtual Knowledge Hub for Business Leaders & Emerging Professionals

AI Prompts to Build This

Copy these build prompts into Claude, Cursor, or your AI coding tool. Create a free account to unlock the full research behind them.

1. 1. Project Setup

   Build the weekend MVP of the "Virtual Knowledge Hub": curated business leaders list office-hours slots, a mentee describes a decision and is matched to three mentors with a reason, and answers can go into a public library. No payment moves through the app this weekend.
   
   Stack: Next.js (App Router, TypeScript), Tailwind, Supabase (Postgres full-text search, Row Level Security, Auth with email magic link), the Anthropic API to re-rank candidate mentors, Resend for booking emails. Mentors paste their own meeting link into each slot. Deploy on Vercel.
   
   Tables (Row Level Security on; mentors read their own rows and public questions, mentees their own rows, the library and approved mentor profiles are public to signed-in users):
   - mentors(id, user_id, display_name, title, company, years_experience, career_arc_text, bio, tags text[], status) status is applied, approved or paused; you approve by hand
   - mentees(id, user_id, display_name, current_role, target_role)
   - questions(id, mentee_id, title, body, desired_outcome, tags text[], privacy, created_at) privacy is public, anonymous or private
   - answers(id, question_id, mentor_id, body, created_at)
   - slots(id, mentor_id, starts_at, duration_min, meeting_url, booked_by, question_id, status) status is open, booked or done
   
   Screens: /login, /ask (describe the moment and see three matches), /mentors/[id], /library (search), /slots (a mentor's slots), /admin/applications.
   Env vars (names only): NEXT_PUBLIC_SUPABASE_URL, NEXT_PUBLIC_SUPABASE_ANON_KEY, SUPABASE_SERVICE_ROLE_KEY (server only), ANTHROPIC_API_KEY, RESEND_API_KEY.
   
   Do not build: billing, plans or mentor payouts, Stripe Connect, embeddings or pgvector, cohorts, video recording, automatic session summaries.
   Done when: npm run dev starts, you can sign in, and the five tables exist with Row Level Security on.

2. 2. Core Feature

   Build the one feature that proves the Hub: a mentee gets three mentors who fit their actual question, and an answer that others can learn from.
   1. A mentor applies with a title, company, years of experience, a short career story and tags. Applications stay in /admin/applications until you approve. Only approved mentors appear anywhere.
   2. A mentee describes their role, the target role, the question, the decision and the outcome they want, and picks a privacy level for the question.
   3. Retrieve candidates in code: Postgres full-text search of the question against mentor profiles plus the overlap of tags. Take the top 20. With fewer than 20 mentors, pass all of them.
   4. Send the question and the candidates to the Anthropic API with a strict JSON schema { rankings: [{ mentor_id, rank, reason }] } asking for the top 3 by how well their actual career story fits this question, with one sentence on why this mentor, not their title. Check that every mentor_id is in the candidate list and drop any that is not.
   5. Show the three mentors with their reason. The mentee books an open slot, which sets booked and emails both people the meeting link the mentor pasted.
   6. A mentor can answer any public or anonymous question in writing. /library searches public questions and answers by full text with filters by tag. An anonymous question shows no name.
   7. A private question is visible only to the mentee and a mentor they booked.
   Rules: never show a private question in the library. Never show a mentor who is not approved. Say that mentors give opinions, not professional advice.
   Empty state: with no approved mentors, tell the mentee the first mentors are being added and offer to email them.
   Done when: an unapproved mentor never shows, a question gets three matches whose ids are all real with a reason each, booking a slot emails both sides, a private question is invisible to others and the library search finds a public answer.

3. 3. Landing Page

   Build a one-page landing site for the Virtual Knowledge Hub, mentorship from seasoned business leaders.
   Hero: "Ask the person who has made your decision before." Sub: "Describe the moment you are in. We match you with three curated mentors whose careers fit it, and you book a short conversation or read the answers in the library." Two buttons: Join the waitlist as a mentee, Apply as a mentor.
   Sections: how it works in three steps (describe the moment, meet your three matches, book or ask), how mentors are curated (reviewed by a person), and an FAQ on cost during the pilot (nothing), what mentors do and do not give (opinions, not professional advice) and how private questions stay private.
   Waitlist: store the email and the role in a waitlist table in Supabase. No other service.
   Style: Geist, a warm off-white background, deep ink type, one navy accent.
   Done when: the page renders on a phone and a submitted email with its role appears in the waitlist table.

4. 4. Branding Package

   Use a design or image tool for this one. A coding agent cannot draw a logo.
   Brand for the Virtual Knowledge Hub: pick a short working name, then design a wordmark and an icon that suggest two speech bubbles of different sizes sharing one tail.
   Colors: warm off-white, deep ink and one navy. Type: Geist, with a serif for mentor stories.
   Deliverables: wordmark, icon, three privacy labels (public, anonymous, private) that differ by shape and word as well as color, a mentor match card with its reason line, and one launch graphic.
   Done when: each deliverable is saved in one folder and the privacy labels are distinguishable without color.