Idea to Implementation Planner

I want to build a SaaS for small businesses that lets them upload invoices and receipts, automatically extracts the important information using AI, and lets them search everything in natural language. I don't want to overcomplicate it, but it needs to be secure because the documents contain financial information. Help me turn this idea into something I can build. Use all relevant .md files in the attached idea-to-implementation-planner package. Apply the skill exactly. Do not invent requirements.

The skill creates a complete, project-specific docs/ folder. These are fully written Markdown documents (.md), not blank templates.

For example:

PRD.md - Product Requirements Document Contains the actual definition of your product: the problem, target users, goals, MVP scope, user journeys, functional and non-functional requirements, assumptions, constraints, success criteria, risks, and open questions.

ARCHITECTURE.md - System Architecture Contains the actual proposed architecture for your product: system overview, architecture diagram, components, responsibilities, data flows, database/storage approach, APIs, authentication and authorization, external services, security boundaries, scalability, reliability, observability, deployment approach, and technology choices.

IMPLEMENTATION_PLAN.md - Implementation Plan Contains the actual step-by-step plan for building your product: development phases, implementation tasks, dependencies, affected components, acceptance criteria, testing requirements, and recommended build order.

ADR/ — Architecture Decision Records Contains individual .md files documenting important technical decisions made for your project, including alternatives considered, why a decision was selected, its trade-offs, consequences, and when it should be reconsidered.

Depending on your project's complexity, the skill may also create:

AI_SYSTEM_SPEC.md - how AI/model components should work, including model choices, prompts, evaluation, guardrails, fallbacks, human review, cost and latency considerations.

SECURITY_AND_COMPLIANCE.md - security requirements, data classification, authentication, authorization, threats, privacy, compliance considerations, and security controls.

RELIABILITY_AND_OPERATIONS.md - reliability targets, monitoring, alerts, deployment, rollback, backups, disaster recovery, incident response, and operational ownership.

TESTING_AND_CICD.md - testing strategy, test coverage, CI/CD, release process, rollback criteria, and definition of done.

RISK_REGISTER.md - identified project risks, their likelihood and impact, mitigations, owners, triggers, and status.

The skill then cross-checks these documents against each other to make sure requirements are reflected in the architecture, architecture decisions are reflected in the implementation plan, and important risks and tests are not missed.

The result is not a collection of templates. It is a project-specific engineering blueprint you can hand to an AI coding agent and use as the foundation for implementation.