Skip to content

A clean, auto-updated prompt library for real-world work: coding, data analysis, engineering, agents & workplace tasks.

License

Notifications You must be signed in to change notification settings

kakarotoncloud/prompt-atlas

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

52 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

🌍 Prompt Atlas

License Last Commit Issues Stars

Prompt Atlas is a clean, auto-updated library of ready-to-use prompts and agent templates for real-world work. It is designed so anyone can use it, including non-technical users, while still being powerful for engineers.

Copy → Paste → Use. No setup required.


✨ What is this?

Prompt Atlas helps you use AI for actual work, not experiments.

You can use it for:

  • Workplace tasks (emails, meetings, planning)
  • Coding and debugging
  • Website building
  • Data analysis
  • Mechanical engineering
  • AI agents and automation

No leaked prompts.
No hidden tricks.
Just clean, reusable prompts.


👥 Who is this for?

🧑‍💼 Non-technical users

You can:

  • write professional emails
  • summarize meetings
  • plan projects
  • improve work quality using AI

You do NOT need coding knowledge.


🧑‍💻 Developers and engineers

You can:

  • debug and refactor code
  • design systems
  • build apps and websites
  • use agent and tool-calling templates

📊 Data analysts

You can:

  • clean datasets
  • perform exploratory data analysis
  • design dashboards
  • extract insights that matter

🔧 Mechanical engineers

You can:

  • perform DFMA reviews
  • do failure analysis
  • structure design calculations
  • plan validation and testing

⚡ Quick start (2 minutes)

Example: Write a professional email

  1. Open prompts/workplace/email-writer.md
  2. Copy the prompt
  3. Paste into ChatGPT / Claude / Gemini
  4. Fill in the placeholders
  5. Done ✅

Example: Debug code

  1. Open prompts/general/debugging.md
  2. Paste your error and code
  3. Get step-by-step debugging help

🧠 How prompts work (simple explanation)

Each prompt:

  • tells the AI what role to act as
  • sets clear rules
  • defines output format
  • leaves placeholders for your input

You only replace the placeholders. Nothing else.


📁 Repository structure (easy guide)

prompts/ ├─ general/ → coding, debugging, system design ├─ workplace/ → emails, meetings, project plans ├─ data/ → data analysis and cleaning ├─ mech/ → mechanical engineering prompts ├─ agents/ → agent and tool-use templates ├─ by-platform/ → ChatGPT / Claude / Gemini wrappers └─ index.md → full prompt list (START HERE) official/ └─ vendor docs → official AI sources (tracked) snapshots/ └─ dated history → weekly change history

New users should start with prompts/index.md.


🤖 Agents and automation (optional)

If you want advanced workflows:

  • multi-step agents
  • tool calling
  • orchestration templates

Look inside prompts/agents/.

This is optional and can be ignored completely.


🔍 Why official sources are included

This repository also tracks official AI documentation pages with:

  • source URL
  • captured timestamp (UTC)
  • SHA256 hash
  • headings outline

These are transparent, auditable, and auto-updated weekly via GitHub Actions.


🧩 Why this pattern and structure is used

This repository intentionally avoids the old “random prompt dump” style.

Most prompt collections:

  • mix unrelated prompts together
  • lack usage context
  • confuse non-technical users
  • become outdated quickly

Prompt Atlas uses a structured, usage-first pattern instead.

Prompts are grouped by real-world tasks:

  • workplace tasks are separated from coding
  • data and engineering prompts have their own space
  • advanced agent workflows are optional, not forced

Each prompt follows a consistent format:

  • clear role definition
  • explicit rules
  • predictable output structure
  • simple placeholders for user input

This makes prompts easier to understand, reuse, and maintain long-term.


🤝 Contributing

You do not need to be an expert.

You can:

  • suggest new prompts
  • improve existing ones
  • report issues

We provide prompt templates and structured issue forms.

See CONTRIBUTING.md.


🛣️ Roadmap

  • more real-world workplace prompts
  • more data and engineering use cases
  • prompt usage examples
  • community contributions

📜 License

MIT License — free to use, share, and adapt.


⭐ Final note

Prompt Atlas is designed to be easy to understand, useful for real work, and maintainable long-term.

If this repo helped you:

  • star it
  • share it
  • suggest improvements

Releases

No releases published

Packages

No packages published

Languages