How to Use The Tools

You’ve got four ways to run the tools on this site:

Paste the Prompt: The simplest method; just drop it into any chat
ChatGPT GPTs: Prebuilt assistants with editable prompts
ChatGPT Projects: Fully custom, model-selectable tool sessions
Kortex AI: Persistent prompts + cross-model support (GPT, Claude, Gemini)

Each has strengths depending on whether you want speed, customisation, reasoning, or cross-platform persistence.

Good for: Instant testing, platform-agnostic use

How it works: Just copy the prompt text from the site and paste it into any chat (ChatGPT, Claude, Gemini etc.)
Control: You decide the model, the context, and the setup
Memory: No persistence unless your platform supports it
Best for: Testing, one-off runs, or use in unsupported platforms

Watch out for:
If the prompt is multi-part or expects a particular structure, you may need to manually separate the setup from the question.

Good for: Fast access and prompt editing in one place

Model: You choose, GPT-5 or GPT-5 Thinking (for explicit reasoning model selection)
Prompt: Editable under the “Configure” tab
History: Conversations go into your ChatGPT history, searchable
@Support: You can use @GPTname in other chats to reference them
Limitations: Doesn't like very large files used for context

Use when: You want a handy, reusable version of a tool without setup fuss.

Good for: Deep thinking tools, prompt adjustments, better output

Model: You choose, GPT-5 or GPT-5 Thinking (for explicit reasoning model selection)
Prompt: Fully visible and editable on start
Conversations: Grouped by Project, not your main chat list
Customisation: High. Edit prompt, tone, model. Easily handles collections of large context files
Search: No @ linking like GPTs, but clean and focused sessions

Use when: You need to add large context files, want to test model quality, or tailor the prompt before diving in.

Good for: Persistent prompt sessions, flexible output saving, and model variety

Models: Supports GPT-4, Claude, Gemini, and more (depending on plan)
Persistence: You paste the prompt once and it stays live for the session

Prompt Management:

• Create a Document titled Prompt: [YourToolName]
• In chat, type @prompt: to open a searchable list and insert it

Storage: Results are only saved if you manually create a document from them
Best for: Persistent tool runs, cross-model testing, clean environments

Use when: You want a streamlined workspace, access to multiple models, or to run tools with precise, repeatable structure.

If you're experimenting beyond ChatGPT, here are a few notes from the field:

Claude (Anthropic)

• Supports direct-paste prompts
• No persistent memory, so you’ll need to re-paste each time
• Markdown formatting not rendered, which means outputs can look plainer
  

That said, Claude has a reflective, sometimes gentle tone that some people prefer. Ideal for exploratory or philosophical prompts.

Gemini (Google)

• Can run prompts via paste
• Tends to stall or over-simplify with multi-perspective or long-form prompts

Works best with shorter, single-role tools or lightweight summaries.

If you like their tone or want to explore alternatives, Kortex AI is worth a look.

It gives you:

• Cross-platform access (GPT, Claude, Gemini, Mistral)
• Prompt persistence, so no need to keep pasting
• A neat workspace with flexible document saving

Just note: the free tier has a limited number of AI queries, so it’s best used for focused sessions rather than casual back-and-forth.

Just trying it out quickly: Paste the Prompt

Bookmark for later, easy re-use: ChatGPT GPT

Prompt editing or large context files: ChatGPT Project

Use with Claude, Gemini etc.: Kortex AI

Cross-session reuse, persistent flow: Kortex AI

Use this step-by-step process to create your own custom GPT, complete with a profile picture, pre-filled name and conversation starters, and the correct model for your prompt.

1. Access Your GPTs
Find the "Explore GPTs button in the left-hand sidebar. At the top, you'll see the “My GPTs” tab.
You can also go directly to:
👉 https://chatgpt.com/gpts/mine

2. Start in the Create Tab
Paste your full prompt into the chat box

The system will:

• Generate a profile picture (you can accept or regenerate)

• Pre-fill the Name, Description, and Conversation Starters

⚠️ Important!
The auto-creation process will soft-interpret your prompt, which can dilute structure or specific instructions.

To fix this:

• Go to the Configure tab
• Clear the Instructions box

Re-paste your full prompt manually

3. Set the Model
Use the Recommended Model dropdown to choose the best model for your prompt:

You can also experiment with o4-mini or o4-mini-high in the dropdown if available.

4. Enable Capabilities
Tick the following in the Capabilities section:

✅ Canvas

✅ Code Interpreter & Data Analysis

5. Finalise & View
To complete the setup:

• Click Create (top-right)
• Choose your sharing option:
• Private/Unlisted
• Public (GPT Store) → Attribution is appreciated!
• Click Save
• Click View GPT to test it live

For most users, especially those running ready-to-use tools, Custom GPTs are the preferred route.

• ✅ No model confusion: Although the model can be changed at runtime, the default is sticky
• ✅ Searchable in chat history: Easy to find later, like a smart app you’ve installed
• ✅ Canvas-enabled: Ideal for writing, reviewing, or visual workflows
• ✅ Friendly front-end: Icon, name, greeting: feels like using a tool, not a dev stub
• ✅ Sharable: Public link, private invite, or store listing: your choice

For general users, GPTs feel clean, discoverable, and ready to use. No techy setup, just start and go.

• In the left sidebar, click New Project.
• Give your project a name.
• Click Add Instructions, paste your prompt, and click Save.
• Optional: change models on the fly, test variations, or edit system prompts live.
  

  Platform Tip: Projects are ideal for fast iteration, testing different models, and fine-tuning tools before turning them into full GPTs.

• Document-first interface: Perfect for longform thinking or layered ideas
• Prompt-as-context: Elegant prompt injection without system prompt juggling
• Great for creators: but markdown/emoji structure may be simplified

Kortex is your studio. Less about public tools, more about building, drafting, and clarity.