Key Concepts

This guide explains the key concepts in PromptBlocks. Once you understand these building blocks, you'll be able to get the most out of the app.

Prefer learning by doing? Follow our Step-by-Step Guide to set up your account with hands-on examples for your role.

The Big Picture

PromptBlocks builds a context layer that makes every AI conversation smarter. Here are the key pieces:

1. Prompts — Reusable text templates you save and use again
2. Global Context — Personal info the AI knows in every conversation
3. Projects — Workspaces with their own context, files, and settings
4. Memories — Facts the AI learns and remembers over time
5. Compositions — Workflows that combine multiple prompts
6. Chat — Where you talk to AI using all of the above

Let's look at each one.

Prompts

A prompt is a piece of text you want to reuse. Instead of typing the same instructions every time, you save them once and reference them later.

Example Prompt

You are a helpful writing assistant. Review the following text
and suggest improvements for clarity, grammar, and flow.

Save this as "Writing Review" in your library, and you can insert it into any conversation by typing @Writing Review.

Variables

Prompts can have variables — placeholders that you fill in when you use them. Variables use double curly braces:

Write a {{tone}} email about {{subject}} to {{recipient}}.

When you use this prompt, PromptBlocks asks you to fill in tone, subject, and recipient. This makes one prompt work for many situations.

Using Prompts

There are two ways to use prompts:

1. In chat: Type @ followed by the prompt name. It inserts the prompt into your message.
2. From the library: Click the copy button to copy a prompt to your clipboard, then paste it anywhere.

Learn more about prompts →

Global Context

Global Context is personal information that gets included in every chat automatically. Set it once, and the AI always knows who you are.

What to Include

• Your role or profession
• Current goals or major projects
• Communication preferences (concise vs. detailed, formal vs. casual)
• Technical background or expertise level

Example

I'm an independent consultant working with mid-size companies on
operational efficiency. I prefer structured thinking: frameworks,
bullet points, clear recommendations.

Current focus: Scaling my practice from 3 to 6 active clients.

With this set, every conversation starts with the AI understanding your context.

Set up Global Context in Settings →

Compositions

A composition combines prompts, text, and dynamic values into a reusable workflow.

What Goes in a Composition?

• Static text — Regular text you type
• Prompts — Your saved prompts, inserted with @
• Meta prompts — Dynamic values like today's date or the current time

Example: Daily Check-in

Good morning! Today is [META: Current Date].

@Daily Planning Prompt

Here are my top priorities:
1. {{priority_1}}
2. {{priority_2}}
3. {{priority_3}}

This composition:

• Automatically inserts today's date
• Pulls in your "Daily Planning Prompt"
• Asks for your priorities when you use it

Meta Prompts

Meta prompts insert dynamic information automatically:

Learn more about compositions →

Projects

Projects help you organize related conversations. Think of them as folders for your AI work.

Why Use Projects?

• Stay organized — Group chats about the same topic
• Add context — Write a project description that the AI sees in every chat
• Attach files — Keep reference documents with your project
• Set defaults — Choose a preferred AI model or assistant for the project

Example

Create a project called "Blog Redesign" with:

• A description of what you're building
• Attached files like brand guidelines or mockups
• All your related conversations in one place

When you chat inside this project, the AI already knows the context.

Learn more about projects →

Memories

Memories are facts the AI learns from your conversations and remembers across chats within a project.

How They Work

• The AI suggests memories based on what you discuss
• You can also tell the AI to "remember that..." explicitly
• Memories are included in future conversations automatically
• Each project has its own memories (up to 50)

What Gets Remembered

• Preferences: "I prefer concise responses"
• Decisions: "We chose the subscription pricing model"
• Key facts: "Budget is $50,000"
• Context: "Sarah is the final decision-maker"

Over time, your projects accumulate knowledge. The AI gets smarter about your work without you having to repeat yourself.

Learn more about personalization →

Tags

Tags are labels you add to prompts for easy filtering. A prompt can have multiple tags.

Examples: #writing, #email, #daily, #work

Use tags to quickly find what you need: filter your library to show only prompts tagged #email when you're writing emails. Tags are flexible—one prompt can belong to multiple categories.

Assistants and Styles

Assistants

An assistant is a custom personality you create for the AI. You write a system prompt that shapes how the AI responds.

Examples:

• "Technical Writer" — Responds with clear, structured documentation
• "Creative Partner" — Brainstorms freely and suggests wild ideas
• "Code Reviewer" — Focuses on code quality and best practices

Communication Styles

Styles adjust the AI's tone without changing its role:

• "Concise" — Short, to-the-point answers
• "Detailed" — Comprehensive explanations
• "Friendly" — Casual and approachable

You can combine assistants and styles. Use your "Technical Writer" assistant with a "Concise" style for brief documentation.

Learn more about settings →

Quick Reference

What's Next?

Ready to set things up? Follow our Step-by-Step Guide for a hands-on walkthrough with examples for your role.

Or explore specific features:

• Chat Interface — Start talking to AI
• Prompt Library — Build your collection
• Compositions — Create workflows
• Projects — Organize your work
• Personalization — How the AI learns about you