Reference

Prompt Doctor, Full Documentation

Investor-, partner- and authority-ready overview of the product, its identity, problem, audience, features, pricing, architecture and business model.

Download as Word document
Version 2.0 · September 2026 · MaxyblaqEasyLife International Limited
Download .docx

Executive Summary

Prompt Doctor is a web-based AI prompt optimisation, education and marketplace platform. It analyses prompts written by humans for Large Language Models (LLMs) such as ChatGPT, Claude, Gemini, Grok and Deepseek, scores their quality, rewrites them at a higher level, and teaches users how to think, not just prompt , better. It also runs a community marketplace where curated, ready-to-use expert prompts can be bought and sold, and a 1:1 consultation service.

  • Legal/operator entity: MaxyblaqEasyLife International Limited (company number 17328188)
  • Registered office: 85 King Richard Street, Coventry, West Midlands, United Kingdom, CV2 4FX
  • Primary domain: https://promptdoctor.co.uk
  • Enquiries: enquiries@promptdoctor.co.uk (email only, no telephone line)
  • Category: AI Tools · Productivity · Education · Marketplace
  • Pricing tiers: Free · Pro (£2.50/month) · Premium (£5/month)
  • Payments: Processed securely by Stripe · seller payout threshold £1.00 GBP
  • Hosting: TanStack Start v1 + Vite 7 on Cloudflare Workers (edge)
  • Backend: Supabase (Postgres, Auth, Storage, Edge Functions)
  • Agent access: Public MCP server at /mcp

1. Product Identity

1.1 Name

Prompt Doctor. The name positions the product as a clinical, expert "diagnostician" for AI prompts: it examines a prompt, identifies thinking flaws, prescribes a remedy (a rewrite), and educates the patient (the user) so the condition doesn't recur. Variants used across the codebase and SEO metadata: PromptDoctor, Prompt Doctor AI.

1.2 Logo

The logo is a caduceus-inspired mark, a circle housing the two intertwined snakes of medical iconography, with the letters "P" and "d" tucked inside the circle, one on either side of the snakes. The medical caduceus reinforces the "doctor" metaphor: this is a tool that heals broken prompts.

1.3 Tagline

A second brain for your prompts. Prompt Doctor diagnoses the thinking flaws behind weak prompts, rewrites them at a higher level, and teaches you to think, not just prompt, better.

1.4 Colour System

The visual identity is built around a deep green + warm gold gradient, with purple used sparingly as an accent and red reserved for action buttons. Body copy is bright white on dark backgrounds and near-black on light backgrounds for maximum contrast. Colours are defined as semantic tokens using oklch() for perceptual consistency across light and dark mode.

TokenLightDarkPurpose
--brand-greenoklch(0.42 0.11 155)oklch(0.65 0.16 155)Primary brand anchor
--brand-goldoklch(0.72 0.15 85)oklch(0.80 0.15 88)Second half of brand gradient
--brand-purpleoklch(0.55 0.20 295)oklch(0.70 0.20 295)Sparing accent only
--destructiveoklch(0.55 0.22 27)oklch(0.62 0.21 27)Action buttons (red)
--prooklch(0.80 0.15 88)oklch(0.82 0.15 88)Pro tier (warm gold)
--backgroundoklch(0.97 0.008 150)oklch(0.17 0.03 155)Page background
--foregroundoklch(0.05 0.01 150)oklch(1 0 0)Body text (high contrast)

1.5 Typography

Default font family is the system UI stack used by Tailwind CSS v4 with semantic font sizing. Headings use tightened tracking; body copy uses relaxed line-height for legibility. Light/dark mode is user-toggleable and persisted.

2. The Problem We Solve

Generative AI tools are now used by hundreds of millions of people across every profession, yet the vast majority of users:

  1. Don't know what a "good" prompt looks like. They type questions the way they'd Google, and receive vague answers.
  2. Blame the AI for their own input. Frustration with "AI is useless" almost always traces to under-specified prompts.
  3. Waste time re-prompting. A single high-quality prompt can replace ten back-and-forth attempts.
  4. Never learn the underlying skill. Most tools improve a single output; they don't teach the user to do it themselves next time.
  5. Can't reliably reuse what worked. Good prompts live in scattered notes, screenshots and DMs.

Prompt Doctor solves all five. It scores prompts on objective rubrics (clarity, context, output format, constraints, role, examples), rewrites them at a higher level, explains why each change matters, and lets users save, share, sell and resell what works. The Academy turns one-off improvements into a permanent skill upgrade.

3. Target Audience

SegmentUse case
Knowledge workers (analysts, consultants, marketers, lawyers, HR, ops)Daily AI co-pilot for research, drafts, summaries, without becoming "prompt engineers".
Solo founders & small teamsOne operator running marketing, sales and support gets senior-level output from junior tools.
Content creatorsUse prompts as production assets, buy from the marketplace and sell what works.
Students & lifelong learnersUse the Academy to learn prompt thinking as a foundational skill.
Developers & technical buildersUse the analyser and marketplace categories for coding, architecture and design.
Prompt creators & sellersMonetise their best prompts with Stripe-backed payouts.
Enterprises & teams (planned)Bulk seats and private prompt libraries.

4. Core Features

4.1 Prompt Analyser /analyze

Paste any prompt and receive a quality score with sub-scores, a diagnosis of thinking flaws, a one-click rewrite tuned to the chosen LLM, voice input + image OCR support, saveable history and a public share link.

4.2 Prompt Generator /doctor

Describe the goal; Prompt Doctor builds a full structured prompt from scratch.

4.3 Personas /personas

Reusable role + context presets that instantly raise prompt baseline quality.

4.4 System Prompts Library /system-prompts

A curated library of 10,000+ professionally written system prompts across every major profession, with server-side pagination.

4.5 Academy /academy

Structured learning path teaching the thinking behind good prompts, not tricks.

4.6 Marketplace /marketplace

Two-sided marketplace with browse, categories, featured, top sellers, new arrivals and a free tier. Sellers receive Stripe payouts.

4.7 Sell Hub /sell

Onboarding for prompt creators: overview, requirements, payouts, pricing tips, FAQ.

4.8 Consultation /consultation

Paid 1:1 prompt-engineering sessions with integrated booking and manage-booking links.

4.9 Dashboard /dashboard

Logged-in home with streak, usage, subscription, quick actions and invite-friends card.

4.10 Search /search

Global, device-responsive live search across pages and the system-prompts catalogue, with debounced URL sync and instant clearing on every form factor.

4.11 Guided Tour /tour

A 14-step guided walkthrough of the whole product, linked from the Product menu, so new users can find every area without hunting.

4.12 Explore & Features /explore, /features

Public overview pages listing every destination in the app, keeping the home page short while still exposing the full site.

4.13 Notifications

Realtime in-app notifications with toast alerts, unread counts, mark-all-as-read and per-category preferences (Updates, Tips, Offers). Admins broadcast announcements from /admin/notifications.

4.14 Floating AI Assistant

Site-wide assistant with translation and text-to-speech.

4.15 Agent integrations (MCP) /mcp

A public Model Context Protocol server lets AI agents browse the marketplace programmatically via the browse_marketplace tool.

4.16 Documentation /documentation

This document, hosted in-app and downloadable as a Word file.

4.17 Help, FAQ, Legal & Contact

Help hub with category sub-routes, contact form, categorised FAQ and a full legal centre (terms, privacy, cookies, disclaimer, accessibility, impressum, AUP, ownership notice, release notes). Contact is by email only at enquiries@promptdoctor.co.uk; no telephone number is collected or published.

4.18 Navigation

One consistent menu on every device, grouped into Product, Pricing, Learn, Marketplace, Sell, Support and Company, with nested sub-menus. Layouts run full width on desktop.

5. Pricing

PlanPriceForIncludes
Free£0Trying it out3 diagnoses + 3 generations / day, browse marketplace, Academy intros, short history
Pro£2.50 / monthDaily AI users10 diagnoses + 10 generations / day, full rewrites, all personas, full Academy, full history, priority queue
Premium£5 / monthPower users & sellers15 diagnoses + 15 generations / day on the top-tier model · everything in Pro · selling rights · payouts · consultation discount

A launch promo offers a 3-month introductory rate. Cancel in one click. Buyers do not need a subscription, only sellers do.

6. Technical Architecture

6.1 Stack

  • Frontend & SSR: TanStack Start v1, React 19, Vite 7, file-based routing.
  • Styling: Tailwind CSS v4 with semantic design tokens; shadcn/ui.
  • Runtime: Cloudflare Workers (edge) via Wrangler.
  • Backend (Lovable Cloud): Supabase, Postgres, Auth (email + Google OAuth), Storage, Edge Functions (Deno).
  • Payments: Stripe, Checkout, embedded checkout dialog, signed webhooks.
  • AI inference: Lovable AI Gateway plus user-configurable external LLMs.
  • Marketing video: Remotion-based explainer videos with narrated, loudness-normalised audio.
  • Email: Transactional and auth emails (sign-up welcome, new-login alerts) from notify.promptdoctor.co.uk.
  • Agent access: Public MCP server exposing marketplace browsing to AI agents.

6.2 Server vs edge functions

App-internal logic uses TanStack createServerFn; AI inference, webhooks, OCR, transcription, TTS and translation run as Supabase Edge Functions; public webhooks live under /api/public/* with HMAC signature verification.

6.3 Security

Row-Level Security on every public table, with explicit GRANTs. Roles live in a dedicated user_roles table with a SECURITY DEFINER has_role() function. Prompt and template content columns are unreadable directly; entitlement-checking functions are the only access path. Helper functions have pinned search_path, shared-analysis links are authenticated, ad impressions are deduplicated and click events require a prior impression, and consultation bookings are written only through trusted server functions. Admin endpoints are gated by auth middleware plus role checks; service-role keys never reach the browser; Stripe webhooks are signature-verified before any state change.

6.4 Authentication

Email/password plus Google sign-in. Route guards resolve the session asynchronously and Google sign-in returns through /auth/callback, so protected pages no longer bounce users back to the sign-in screen.

6.5 SEO & discoverability

Unique title/meta/OG per route, canonical URLs, JSON-LD SoftwareApplication schema, server-side /sitemap.xml covering all public routes, and llms.txt for AI crawlers. robots.txt blocks only admin, API and internal paths; private, auth-gated routes carry noindex meta tags instead.

6.6 Advertising compliance

AdSense script in the document head on every page, Google Consent Mode v2 with a cookie consent banner, ads.txt, ad units marked up semantically and clearly labelled, and no ad placement adjacent to navigation.

6.7 Accessibility

Keyboard-focusable nav with live ARIA announcements, prefers-reduced-motion respected, high-contrast body text in both themes (bright white on dark, near-black on light), and a dedicated accessibility statement.

7. Business Model & Revenue Streams

  1. Subscriptions, Pro (£2.50) and Premium (£5) recurring monthly revenue.
  2. Marketplace commission, Platform fee on each sale; seller keeps the majority. Payouts are released once a seller's balance reaches the £1.00 GBP minimum threshold.
  3. 1:1 Consultations, Paid bookings.
  4. Advertising, Non-intrusive AdSense on free-tier pages.
  5. Future, Team seats, white-label libraries, public API access.

8. Competitive Differentiation

Most "prompt" tools are either directories with no analysis, single-shot rewriters with no education, or developer SDKs that only help engineers. Prompt Doctor combines all four: analyse + rewrite with explainability, generate from scratch, teach the skill through the Academy, and monetise expertise through a two-sided marketplace. A clinical brand identity, aggressive £2.50 entry pricing and multi-LLM support give it a defensible niche between consumer AI tools and enterprise prompt-ops platforms.

9. Roadmap (next 12 months)

  • Team workspaces with shared prompt libraries.
  • Public API for the analyser (B2B).
  • iOS & Android wrappers over the responsive web app.
  • Expanded Academy with certifications.
  • Localisation into the top 10 LLM-using languages.
  • Affiliate / referral programme.
  • Enterprise SSO via Supabase SAML.

10. Investment Highlights

  • Live, revenue-generating product with Stripe-validated paid plans from day one.
  • Low unit cost, edge-hosted, AI Gateway pricing model, no GPU fleet to operate.
  • Multiple revenue streams (subs, marketplace, consultations, ads, future API).
  • Defensible content moat, 10,000+ curated system prompts and a growing Academy.
  • Modern, scalable stack, TanStack Start + Cloudflare Workers + Supabase.
  • Brand-first identity, memorable name, distinctive caduceus logo, clear positioning.
  • Built-in virality, shareable analyses, marketplace listings and a referral programme.

11. Contact

Prefer a portable copy? Download the full document as Word.
Download .docx