Note: This project is a curated collection of AI-driven development prompts and workflows, adapted from and inspired by the BMAD Method (Business-Managed Agile Development). It is a standalone resource designed to accelerate software delivery.
This repository serves as a "Prompt Engineering Toolkit" for the entire Software Development Life Cycle (SDLC). It contains a structured library of highly specialized prompts that turn any LLM (Gemini, Claude, GPT) into an expert role-player for specific sprint tasks—from initial market research to final release documentation.
Instead of writing generic requests, we use these "Artifacts" to ensure consistent, high-quality, and standard-compliant outputs at every phase.
The core of this project is the docs/sprint-artifacts/ directory, organized by development phase:
01-planning-strategy/: (The "Product Owner" Phase)- Generates PRDs, User Stories, Competitive Analysis, and Market Research.
02-design-ux/: (The "Designer" Phase)- Conducts UX Audits, Accessibility Checks, and Design System reviews.
03-engineering-audit/: (The "Architect" Phase)- Analyzes Codebase Health, Security Risks, Database Schemas, and Infrastructure.
04-implementation/: (The "Builder" Phase)- Provides rigorous prompts for writing clean, type-safe Frontend, Backend, and Fullstack code.
05-testing-qa/: (The "Tester" Phase)- Generates comprehensive Test Plans, Unit Tests, and Visual QA strategies.
06-documentation/: (The "Tech Writer" Phase)- Automates the creation of READMEs, API References, and Release Notes.
We follow a structured flow to ensure no context is lost between phases.
👉 See the Artifact Workflow Map for a visual guide on the optimal sequence of prompts.
- Identify your current phase in the SDLC (e.g., "I have a vague idea" -> Phase 1, or "I have code that needs testing" -> Phase 5).
- Locate the relevant prompt file.
- Copy & Paste the prompt content into your AI chat.
- Append your context (e.g., your code snippets, rough notes, or previous artifact outputs).
This workflow methodology is heavily influenced by the BMAD Method, adapting its agent-based architecture into a lightweight, prompt-focused toolkit for agile teams.