skills.rest
Deepak-ai-93
Deepak-ai-93COMMUNITY
Explore all skills by this author →
GitHub

prompt-engineering

Builds a versioned prompt library in the author's voice with framework checks and test verdicts.

1|Updated Aug 7, 2026
Search Tags:#prompt engineering#audit#prompt library#content workflow#ai prompts#prompt testing#voice rules
One-click install
npx skills add https://github.com/Deepak-ai-93/deepak-skill --skill prompt-engineering-deepak-ai-93
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: prompt-engineering
Source: https://github.com/Deepak-ai-93/deepak-skill/tree/main/skills/prompt-engineering
Command: npx skills add https://github.com/Deepak-ai-93/deepak-skill --skill prompt-engineering-deepak-ai-93

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve? Reusable AI prompts often sound generic, lack structure, and are never tested, so outputs drift from the creator's voice and quality is inconsistent. This Skill turns a creator's recurring AI use cases into a structured, versioned prompt library where every prompt follows a fixed framework, enforces the author's voice rules, and carries an honest test verdict. ## Core Features & Use Cases - Five-part prompt framework: Every prompt is written with Role, Context, Task, Format, and Constraints, with no placeholder gaps, so each prompt is copy-paste ready. - Voice rules from a taste profile: A memory file captures the author's tone, rhythm, and banned words, which become the library's Voice rules section referenced by every prompt. - Test and audit loop: Each prompt is run against a real input with a pass/needs-work verdict recorded, then an automated audit script plus a fresh auditor subagent score the library out of 50 before delivery. - Use Case: A B2B content creator asks for a prompt library covering hooks, newsletter intros, competitor teardowns, and carousel repurposing; the Skill delivers prompt-library.md, test-results.md, and a signed prompts-audit.md. ## Quick Start Build me a prompt library for my weekly content workflow covering hooks, newsletters, and research, written in my voice and tested before delivery.

Frequently Asked Questions about prompt-engineering

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I build a reusable prompt library for my content workflow?▼

Identify 3-6 use cases you actually run weekly, then write each prompt with Role, Context, Task, Format, and Constraints. Test every prompt on a real input, record a pass or needs-work verdict, and rewrite failures before shipping the library.

How do I make AI prompts write in my own voice?▼

Capture a taste profile with your tone, sentence rhythm, and banned words, then place it as a Voice rules section at the top of the library. Every prompt's constraints must reference those rules so outputs match your style instead of generic AI phrasing.

What structure should a good AI prompt have?▼

A strong prompt has five parts: Role defining who the AI is, Context with what it needs to know, Task naming specific inputs and angles, Format defining the exact output shape, and Constraints with measurable caps and banned words.

Does this prompt library work with ChatGPT and Claude?▼

Yes, prompts are written to be portable across tools like Claude, ChatGPT, and Gemini, and each prompt records which tool it was tested with. The test results file notes the tool per prompt so you can match prompts to the right model.

Why do my AI prompts produce generic output?▼

Generic output usually comes from vague tasks and missing constraints, such as asking to write a hook without angles, caps, or voice rules. Adding specific inputs, character limits, banned words, and a voice rules reference makes outputs consistent and on-brand.

How are prompts tested before the library is delivered?▼

Each prompt runs against one real sample input and gets a quality score plus a pass or needs-work verdict in test-results.md. An automated audit script then checks framework completeness and placeholders, and a fresh auditor subagent scores the library out of 50.

Related Skills

Explore More Content & Communication Skills→
warpiratewarpirate
4

train-voice

Analyze writing samples to generate a reusable markdown voice profile.

Community
Intermediate
ntd237ntd237
1

content-adapt-multichannel

Convert long-form content into TikTok scripts, carousel slides, and email newsletters.

Community
Intermediate
mohamednegm0mohamednegm0

brand-building

Define brand strategy, positioning, messaging, and voice guidelines for B2B and product brands.

Community
Intermediate
AO-Cyber-SystemsAO-Cyber-Systems

df:quick

Execute ad-hoc tasks with atomic commits and STATE.md updates.

Official
Intermediate