How does the skill handle existing project context?

It performs 'Step 0: Repository Context' where it reads your README, documentation, and entry points to identify existing conventions and dependencies before suggesting a plan.

Where are the generated requirements stored?

All PRDs are saved as Markdown files in a standardized `tasks/` directory within your repository, using a `prd-[feature-name].md` naming convention.

Does this skill start writing code automatically?

No, the skill explicitly focuses on the planning phase. It stops after generating the PRD to ensure the user agrees with the scope before any implementation begins.

What makes these PRDs 'AI-ready'?

These PRDs use 'Task Definitions'—single, detailed descriptions of work—and verifiable 'Acceptance Criteria' that allow AI agents to validate their own work without needing human intervention for clarification.

AI-Ready PRD Generator

Name: AI-Ready PRD Generator
Author: frizynn

byfrizynn

•

生产力与工作流

Generates detailed, agent-executable Product Requirements Documents (PRDs) by analyzing repository context and clarifying feature goals through interactive questioning.

The PRD skill is designed to bridge the gap between high-level feature ideas and executable code by producing highly structured Product Requirements Documents. It specifically optimizes documentation for AI agents and junior developers, ensuring that every task is unambiguous and verifiable. By first gathering repository context and then guiding the user through a series of targeted, multiple-choice clarifying questions, it eliminates scope creep and architectural misalignment. The resulting markdown files include precise user stories, technical constraints, and success metrics, providing a single source of truth that is ready for immediate implementation in tools like Claude Code, Cursor, or Amp.

主要功能

01Explicit scope boundary setting through dedicated Non-Goals sections

02Automated repository context analysis to ensure architectural alignment

03Interactive multiple-choice questioning to resolve feature ambiguity

04Structured PRD output with actionable Task Definitions and Acceptance Criteria

0515 GitHub stars

06Automatic file management with standardized naming in the tasks/ directory

使用场景

01Generating granular task lists for autonomous AI agents to execute independently

02Scoping out new features for complex codebases where context is critical

03Standardizing the requirement gathering process for engineering teams using Git worktrees

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add frizynn/gralph prd

For use in Claude.ai and ChatGPT

Download Skill