The Ultimate Foundation for AI-Powered Creativity
A curated collection of expertly crafted prompts and agent skills for every AI use case
In the rapidly evolving world of AI, the right prompt can unlock extraordinary results. This repository serves as your definitive resource for:
- High-quality prompts tested across multiple AI platforms
- Agent Skills following the Agent Skills specification and discoverable via skills.sh — install with
npx skills add pokanop/aior as a Claude Code plugin - Best practices distilled from real-world usage
- Foundation patterns applicable to any AI use case
Whether you're generating stunning visuals, crafting compelling videos, synthesizing audio, or building intelligent agents — this is your toolkit.
Plenty of collections offer prompts or agent skills. This one is built on three commitments that most don't make:
- Artifact-driven. Every lifecycle skill emits real, reviewable filesystem artifacts in
plans/— PRDs, designs, ADRs, task lists, decision logs, retros. The work leaves a paper trail you can read, diff, and audit, not just a chat transcript. - Auditable. Deterministic plan validation (
plan-validate.py,plan-metrics.py) checks task lists mechanically instead of by eye, shared verification gates require evidence before any completion claim, and an eval harness plus CI checks keep the skills themselves honest. - Multimodal. The SDLC skills live alongside curated image, video, audio, and text prompt libraries — one repository, one docs site, covering both how agents build software and how they generate media.
Prompts optimized for:
|
Master prompts for:
|
Prompts for:
|
Expert prompts for:
|
Skills in this repository follow the Agent Skills Specification — an open format for giving AI agents new capabilities and expertise. Skills are discoverable and installable via the skills.sh registry.
Agent Skills are folders containing instructions, scripts, and resources that agents can discover and use to perform tasks more accurately and efficiently. Each skill is a self-contained directory with a SKILL.md file.
| Skill | Description |
|---|---|
code-review |
Structured code review process with actionable feedback |
debug-and-fix |
Systematic debugging and root-cause analysis |
dependency-upgrade |
Safe, staged dependency and major-version upgrades |
design-to-tasks |
Break a design (or PRD) into actionable engineering tasks |
idea-to-prd |
Create a Product Requirements Document from a feature idea |
incident-postmortem |
Blameless incident postmortem with routed action items |
next-step |
Report plan status and route to the right skill |
performance-review |
Measurement-driven whole-system performance sweep |
plan-retrospective |
Reflect on completed work and capture learnings |
prd-to-design |
Turn a PRD into a technical design and ADRs |
refactor |
Restructure code safely without changing behavior |
release-checklist |
Pre-release verification and deployment checklist |
security-review |
Lightweight threat model and security posture sweep |
setup-pokanop-skills |
One-time repo onboarding for the skill suite |
tasks-to-code |
Implement engineering tasks from a task list |
ui-design-audit |
Audit UI components for design consistency |
write-tests |
Add or backfill test coverage without changing behavior |
See the Skills README for complete documentation, recommended order of operations, and workflow details.
Each skill lives in its own directory inside skills/:
skills/
└── skill-name/
├── SKILL.md # Required: main instructions + frontmatter
├── scripts/ # Optional: executable helper scripts
├── references/ # Optional: additional documentation
└── assets/ # Optional: static resources
Every SKILL.md must include YAML frontmatter with at minimum name and description:
---
name: skill-name
description: >
What this skill does and when to use it. Include specific keywords
that help agents identify when the skill is relevant.
license: MIT
metadata:
author: pokanop
version: "1.0"
---
# Skill Instructions
Detailed instructions and guidance for the agent...| Field | Required | Description |
|---|---|---|
name |
✅ Yes | Lowercase alphanumeric + hyphens only. Must match the directory name. |
description |
✅ Yes | What the skill does and when to use it (1–1024 chars). |
license |
No | License name or reference (e.g. MIT, Apache-2.0). |
compatibility |
No | Environment requirements, target agent products, or required tools. |
metadata |
No | A map of arbitrary string keys/values (e.g. author, version). |
allowed-tools |
No | List of tools the skill is permitted to use. |
namerules: 1–64 characters, lowercasea-zand-only, must not start or end with-, no consecutive--. Must match the parent directory name exactly.
- Create a directory in
skills/using lowercase letters and hyphens (e.g.skills/my-skill/) - Add a
SKILL.mdwith valid frontmatter and instructions - Optionally include
scripts/,references/, orassets/subdirectories
Install any skill from this repo via the CLI:
npx skills add pokanop/aiSupported by: Claude Code, Cursor, Windsurf, OpenCode, GitHub Copilot, OpenHands, and more.
See the Agent Skills Specification for the complete format reference.
ai/
├── assets/ # SVGs, images, and visual resources
│ ├── logo.svg
│ ├── hero-banner.svg
│ ├── features-icons.svg
│ ├── icon-video.svg
│ ├── icon-audio.svg
│ ├── icon-agents.svg
│ └── icon-code.svg
├── prompts/
│ └── images/ # Image generation prompts (videos, audio, text coming soon)
├── skills/ # Agent Skills (agentskills.io format)
│ ├── code-review/
│ ├── debug-and-fix/
│ ├── design-to-tasks/
│ ├── idea-to-prd/
│ ├── next-step/
│ ├── performance-review/
│ ├── plan-retrospective/
│ ├── prd-to-design/
│ ├── refactor/
│ ├── release-checklist/
│ ├── security-review/
│ ├── tasks-to-code/
│ └── ui-design-audit/
└── README.md
# Clone the repository
git clone https://github.com/pokanop/ai.git
# Navigate to your use case
cd ai/prompts/imagesInstall all skills in this repository directly into your project:
npx skills add pokanop/aiOr browse and install individual skills from the skills.sh leaderboard.
This repository is designed to be the foundation for all AI use cases:
- Generate consistent brand imagery
- Create video content at scale
- Produce original music and soundtracks
- Write engaging content effortlessly
- Build intelligent agents with specialized skills
- Automate repetitive coding tasks
- Create AI-powered workflows
- Integrate prompts into applications
- Standardize AI interactions across teams
- Maintain brand voice consistency
- Scale content production
- Reduce AI iteration time
- Study prompt engineering patterns
- Benchmark AI model performance
- Contribute to collective knowledge
- Share reproducible experiments
[CONTEXT] - Who/what is involved
[TASK] - What needs to be done
[CONSTRAINTS] - Limitations and requirements
[STYLE] - Tone, format, aesthetic
[EXAMPLES] - Reference outputs (optional)
| Category | Key Tips |
|---|---|
| Images | Be specific about style, lighting, composition |
| Videos | Describe motion, transitions, timing |
| Audio | Specify mood, tempo, instruments |
| Text | Define audience, tone, length |
| Agents | Set clear goals, constraints, success criteria |
We welcome contributions! Here's how to help:
- Add prompts - Share your best-performing prompts
- Create skills - Build reusable agent capabilities following the specification
- Improve docs - Enhance explanations and examples
- Report issues - Found something that doesn't work? Let us know
See CONTRIBUTING.md for detailed guidelines.
- Interactive prompt playground
- Prompt effectiveness ratings
- Multi-language support
- AI model compatibility matrix
- Community prompt sharing
- Video tutorials
This project is licensed under the MIT License - see the LICENSE file for details.
Made with care for the AI community
Star this repo if you find it useful!
Report Bug · Request Feature · Join Discussion
Building the future of AI, one prompt at a time
If this project helped you, consider supporting its development:
