Awesome/awesome-copilot
1
0
Fork 0
mirror of https://github.com/github/awesome-copilot.git synced 2026-02-20 02:15:12 +00:00
Code Issues Packages Projects Releases Wiki Activity

Simplify PRD skill by removing technical jargon

Browse Source 
Remove subagent, model selection, and implementation details that users don't need to specify. Focus on business requirements and user needs.
This commit is contained in:
Aung Myo Kyaw
2026-01-23 00:36:52 +07:00
parent e24be77e6f
commit 03986ae177
1 changed files with 13 additions and 20 deletions
Download Patch File Download Diff File Expand all files Collapse all files

33
skills/prd/SKILL.md
View File

@@ -1,6 +1,6 @@
--- ---
name: prd name: prd
description: 'Generate high-quality Product Requirements Documents (PRDs) for AI agents and software systems. Includes executive summaries, user stories, technical specifications, AI-specific requirements (models, data, prompts), and risk analysis.' description: 'Generate high-quality Product Requirements Documents (PRDs) for software systems and AI-powered features. Includes executive summaries, user stories, technical specifications, and risk analysis.'
license: MIT license: MIT
--- ---
@@ -8,7 +8,7 @@ license: MIT
## Overview ## Overview
Design comprehensive, production-grade Product Requirements Documents (PRDs) that bridge the gap between business vision and technical execution. This skill is optimized for AI agent development and modern software systems, ensuring that both deterministic and non-deterministic requirements are clearly defined. Design comprehensive, production-grade Product Requirements Documents (PRDs) that bridge the gap between business vision and technical execution. This skill works for modern software systems, ensuring that requirements are clearly defined.
## When to Use ## When to Use
@@ -16,7 +16,7 @@ Use this skill when:
- Starting a new product or feature development cycle - Starting a new product or feature development cycle
- Translating a vague idea into a concrete technical specification - Translating a vague idea into a concrete technical specification
- Defining requirements for an AI agent (LLM, ML, etc.) - Defining requirements for AI-powered features
- Stakeholders need a unified "source of truth" for project scope - Stakeholders need a unified "source of truth" for project scope
- User asks to "write a PRD", "document requirements", or "plan a feature" - User asks to "write a PRD", "document requirements", or "plan a feature"
@@ -33,7 +33,6 @@ Before writing a single line of the PRD, you **MUST** interrogate the user to fi
- **The Core Problem**: Why are we building this now? - **The Core Problem**: Why are we building this now?
- **Success Metrics**: How do we know it worked? - **Success Metrics**: How do we know it worked?
- **Constraints**: Budget, tech stack, or deadline? - **Constraints**: Budget, tech stack, or deadline?
- **Agent Roles**: If this is for an AI system, what are the subagent responsibilities?
### Phase 2: Analysis & Scoping ### Phase 2: Analysis & Scoping
@@ -41,7 +40,6 @@ Synthesize the user's input. Identify dependencies and hidden complexities.
- Map out the **User Flow**. - Map out the **User Flow**.
- Define **Non-Goals** to protect the timeline. - Define **Non-Goals** to protect the timeline.
- Identify **Orchestration Needs** (which tools or agents are required?).
### Phase 3: Technical Drafting ### Phase 3: Technical Drafting
@@ -85,12 +83,10 @@ You **MUST** follow this exact structure for the output:
- **Acceptance Criteria**: Bulleted list of "Done" definitions for each story. - **Acceptance Criteria**: Bulleted list of "Done" definitions for each story.
- **Non-Goals**: What are we NOT building? - **Non-Goals**: What are we NOT building?
### 3. AI & Agent Orchestration (If Applicable) ### 3. AI System Requirements (If Applicable)
- **Model Selection**: e.g., `Claude 3.5 Sonnet` for reasoning, `Haiku` for speed. - **Tool Requirements**: What tools and APIs are needed?
- **Agent Definitions**: Specific roles for subagents (e.g., `librarian` for research). - **Evaluation Strategy**: How to measure output quality and accuracy.
- **Tool Whitelist**: Explicit list of tools each agent is allowed to use.
- **Evaluation Strategy**: How to measure AI output quality (Evals, Golden Sets).
### 4. Technical Specifications ### 4. Technical Specifications
@@ -109,8 +105,7 @@ You **MUST** follow this exact structure for the output:
### DO (Always) ### DO (Always)
- **Delegate Visuals**: If the PRD involves UI/UX, explicitly instruct the use of the `frontend-ui-ux-engineer` agent. - **Define Testing**: For AI systems, specify how to test and validate output quality.
- **Define Evals**: For AI systems, specify the **Evaluation Protocol** (how to detect hallucinations or failures).
- **Iterate**: Present a draft and ask for feedback on specific sections. - **Iterate**: Present a draft and ask for feedback on specific sections.
### DON'T (Avoid) ### DON'T (Avoid)
@@ -120,12 +115,12 @@ You **MUST** follow this exact structure for the output:
--- ---
## Example: AI-Powered Search Agent ## Example: Intelligent Search System
### 1. Executive Summary ### 1. Executive Summary
**Problem**: Users struggle to find specific documentation snippets in massive repositories. **Problem**: Users struggle to find specific documentation snippets in massive repositories.
**Solution**: A RAG-based search agent that provides direct answers with source citations. **Solution**: An intelligent search system that provides direct answers with source citations.
**Success**: **Success**:
- Reduce search time by 50%. - Reduce search time by 50%.
@@ -138,13 +133,11 @@ You **MUST** follow this exact structure for the output:
- Supports multi-turn clarification. - Supports multi-turn clarification.
- Returns code blocks with "Copy" button. - Returns code blocks with "Copy" button.
### 3. Agent Orchestration ### 3. AI System Architecture
- **Primary Agent**: `Oracle` for reasoning and answer synthesis. - **Tools Required**: `codesearch`, `grep`, `webfetch`.
- **Subagent**: `Librarian` for searching docs and indexing code.
- **Tool Whitelist**: `codesearch`, `grep`, `webfetch`.
### 4. Evaluation ### 4. Evaluation
- **Benchmark**: Use a 'Golden Set' of 50 common dev questions. - **Benchmark**: Test with 50 common developer questions.
- **Pass Rate**: 90% must match ground truth citations. - **Pass Rate**: 90% must match expected citations.