AI Product Requirements Workflow
Turn a fuzzy business need into requirements a team can build from — interrogate the need into concrete requirements, shape them as user stories, and write the PRD.
The full path to a two-sided platform — define the buyer-and-seller requirements, model the data, design the API, build roles and permissions, wire integrations, design the UI, then test, secure, and ship it.
A marketplace is two products in a trench coat: a seller experience and a buyer experience, joined by listings, permissions, and the integrations that move money and messages between strangers who don't trust each other yet. That two-sidedness is what separates it from a SaaS app, a CRM, or an internal tool — and it's where the hard decisions live. This project builds it in the order those decisions demand: pin the requirements for both sides, model the data that ties buyers, sellers, and listings together, design the API, build the roles and permissions that keep each side in its lane, wire the third-party integrations, design the UI components both sides share, and ship. It owns the marketplace outcome specifically — not a CRM (internal sales data), not an internal tool (one audience), not a generic SaaS MVP — so every stage serves a platform with two audiences and the trust boundaries between them. Each stage connects to a NewPrompt workflow you can run on its own; together they carry a marketplace from concept to a platform real buyers and sellers can use. You own the marketplace model and the code; the project keeps the two-sided decisions in the right order.
Each stage runs a NewPrompt workflow, with a supporting resource and tool. Work them in order — the output of each stage feeds the next.
See the execution map →Define the two-sided requirements
Pin what both sides need — sellers listing and managing, buyers discovering and transacting — and the rules that govern their interaction. A marketplace built from one side's requirements fails the other.
Model the marketplace data
Design the schema that ties users, listings, and orders together — the relationships and constraints a two-sided platform lives on — and the indexes the search and discovery queries will actually need.
Design the API
Design the API surface both sides build against — listings, search, orders, messaging — as a contract before code, so the seller tools and buyer experience develop against something stable.
Design the UI components
Design the reusable component system both sides share — listings, dashboards, search, checkout — so the seller and buyer experiences stay consistent instead of diverging into two mismatched products.
Build roles and permissions
Design the access control that keeps each side in its lane — buyer, seller, admin — so a seller can't touch another's listings and a buyer can't reach the admin panel. On a marketplace, permissions are the trust model.
Wire the integrations
Connect the third parties a marketplace runs on — payments, notifications, shipping or messaging — via webhooks and events, with the retries and idempotency that keep a payment or order event from being lost or double-processed.
Lock behavior down with tests
A marketplace moves money and listings between strangers — build a test suite that fails for real reasons across the order, payment, and permission paths, so you can keep shipping without breaking a transaction someone depends on.
Run a security review
A marketplace handles money, personal data, and two distrusting sides — review the auth, payment, and input paths the way an attacker would, and back the findings with tests, before real users and real transactions arrive.
Ship the marketplace
Cross from built to live — readiness checked, rollback planned, monitoring in place — because launching a platform that handles other people's money and listings is exactly where a careful release earns its keep.
Validate the product with real evidence
A marketplace only works if both sides show up. Measure real liquidity — listings, matches, GMV — against the success signal you set, render a hit/partial/miss verdict, and decide whether to iterate, pivot, or scale before widening the funnel on either side.
A marketplace platform built on its two-sided reality — requirements for both audiences, a data model tying buyers, sellers, and listings together, an API, roles and permissions that hold the trust boundaries, integrations, a shared UI component system, and the whole thing shipped — a real platform, not a single-audience app with a marketplace label.
Marketplace is two-sided: many sellers, many buyers, payouts, and trust between strangers. For your own single catalog, use the E-commerce Store project — it is a lighter build.
Because two-sided platforms compound complexity: buyer and seller requirements, roles and permissions, integrations, and a security review all interact. The journey orders them so they do not collapse on each other.
Typically yes — marketplaces move money between parties. The integration and security stages cover the stack that makes that safe.
You run it through the blueprint's ten stages in the order a two-sided platform demands: pin buyer-and-seller requirements, model the data tying users, listings, and orders together, design the API, build roles and permissions, wire integrations, design shared UI, then test, security-review, and ship. AI assists each stage; you own the marketplace model and code.
At minimum, both sides working end to end: sellers can list and manage, buyers can discover and transact, and the roles and permissions keep each side in its lane. Skip nothing on the trust boundary. A marketplace where only one audience functions is not an MVP, it is a single-audience app with a label.
You measure real liquidity, not signups. The final validation stage tracks listings, matches, and GMV against a success signal you set, renders a hit, partial, or miss verdict, and drives an iterate, pivot, or scale decision before you widen the funnel on either side. A marketplace only works when both sides show up.
Turn a fuzzy business need into requirements a team can build from — interrogate the need into concrete requirements, shape them as user stories, and write the PRD.
Design a schema on its data, not a hunch — model the entities and relationships, set the constraints that protect integrity, plan indexes around real queries, then document the schema and migration.
Design an API on its contract instead of discovering it endpoint by endpoint — model the resources, design the endpoints and payloads, pin the contract, then review it before code locks it in.
Design access control before you build it, not after a breach — choose the authentication approach, model the roles and permissions, review the design for gaps, then document the identity model.
Connect systems so they don't break each other — map the integration boundaries, design the event and webhook contracts, plan retries and failure handling, then document the integration.
Structure a UI so it stays consistent as it grows — inventory the screens, break them into reusable components, specify the component system and its rules, then review the structure for drift.
Build a test suite that fails for real reasons, not green decoration — coverage across unit, integration, and edge cases, then a review for the gaps.
Review code for what an attacker would do, not just what tests catch — anchor the model as a security engineer, run a threat-focused review, then back the findings with auth and input tests.
Cross the gap between 'tests pass' and 'safe in production' — assess release readiness, plan the deploy and its rollback, and set up the monitoring and launch checks before you ship, not after.
Find out whether the thing you shipped actually worked — define the success metric, plan the measurement, classify the real evidence, then render a verdict and an iterate / pivot / scale decision.
Turn a rough feature idea into a structured requirement: problem statement, acceptance criteria, and what's explicitly out of scope.
Map every predicate, join, and sort to the index that serves it — or doesn't. Composite order rules, covering decisions, and the write tax nobody mentions.
Architecture review for the public surface: abstractions that earn their place, dependency direction, seams, and the patterns the codebase already has.
Make AI act as a security engineer — threat modeling, secure design, risk assessment, and defensive recommendations — instead of a generic assistant.
You have the schema — fields, types, requirements. The translation into a prompt the model actually follows: schema lines, realistic example, and validation rules.
Make AI act as a software engineer — design, implementation, and testing trade-offs — instead of a generic assistant that writes whatever you ask.
Mock the dependencies, test the business logic, one behavior per test — the unit testing contract that bans plumbing tests.
Twelve security checks — injection, auth, secrets, SSRF, privilege escalation — reviewed the way an attacker would read the code.
SHIP or DO NOT SHIP: twelve checks on failure modes, observability, rollback, and load — the review that happens before the incident.
Design a role model that scales — roles derived from real permissions, the role-permission mapping, and the inheritance and edge cases before you hardcode 'isAdmin'.
Make access auditable — a roles-by-actions matrix that shows exactly who can do what, exposes the gaps and over-grants, and becomes the spec the code is checked against.
Keep tenants out of each other's data — design the isolation model, the tenant-scoping rule every query must obey, and the cross-tenant leaks to test for.
Turn a deploy into a deploy you can undo — a step-by-step rollback plan with triggers, the reverse of every forward step, and the point of no return named.
Consolidate the whole validation into one shareable report — the success metric, the evidence, the hit / partial / miss verdict, and the iterate / pivot / scale decision, in a form the team and stakeholders can act on.
Turn product requirements and key entities into an ERD-ready schema spec: tables, fields, primary and foreign keys, relationships with cardinality, indexes, and constraints — with the modeling gaps named as open questions.
Turn UI requirements and component needs into a design-system spec: a component inventory with states, variants, responsive behavior, accessibility requirements, microcopy rules, and developer handoff notes.
Build structured system prompts from role, tone, constraints, and model target.
Build evidence-based SQL optimization prompts — goal, platform, and the evidence you have turn into a query tuning contract.
Build prompts that return structured data — JSON first, with YAML, XML, and CSV modes — parseable every time.
Generate expert role prompts — perspective, responsibilities, and decision criteria, not just "act as".
Break a big goal into a sequential prompt workflow — each step with an objective, prompt, and expected output.
Build prompts that produce documents in a fixed structure — headings, sections, and tables.
Generate code review prompts for any scope — snippet, file, diff, or full pull request — with focused review rules.
Build test generation prompts — unit, integration, or E2E — with framework modes and edge-case coverage rules.
Practical project ideas you can build from this base project path — each opens in the Project Advisor.
The full path to a support agent you can put in front of customers — write its instructions, ground it in your docs, route and handle tickets, then evaluate and cost-control it before it goes live.
The full path from idea to a shipped SaaS MVP — define and scope the requirements, design the architecture, API, and data model, then build it reviewed, tested, secured, cost-controlled, and deployed.
The full path to a backend you can put clients on — define the requirements, design the architecture, API contract, data model, and access control, then build it reviewed, tested, secured, and shipped.
The full path to a CRM that fits your sales process — define the contacts, deals, and pipeline, model the data that ties them together, then build the roles, integrations, and pipeline UI, and ship.