Skip to main content

​
Overview

Memories are knowledge snippets that provide context for your projects. Store notes, guidelines, and information that helps both humans and AI understand your project.

​
What Are Memories?

Memories are:
  • Text-based knowledge snippets
  • Project-scoped contextual information
  • AI-accessible via MCP servers
  • Searchable and manageable

​
Use Cases

Project Notes

Store important notes, decisions, and observations about your project.

AI Context

Provide guidelines and context that AI models should know when accessing your project.

Knowledge Snippets

Capture insights, lessons learned, and key information.

Documentation

Store project documentation, procedures, and references.

​
Creating Memories

1

Navigate to Project

Open the project where you want to add a memory.
2

Go to Memories Tab

Click the “Memories” tab in the project dashboard.
3

Click Add Memory

Click the “Add Memory” button.
4

Fill in Details

A dialog appears with two fields:Name (required):
  • Short, descriptive title
  • Example: “Project Guidelines”, “Research Questions”, “Key Findings”
Content (required):
  • Full text of the memory
  • Can be multiple paragraphs
  • Plain text (not Markdown)
5

Save Memory

Click “Save” or “Create” to save the memory.
Memory appears in the Memories tab list.

​
Memory Content Guidelines

​
What to Include

Good Memory Content:
  • Context about the project
  • Guidelines for AI interactions
  • Key decisions and rationale
  • Important facts and figures
  • Lessons learned
  • References to related information
Example: 
This project analyzes customer feedback from Q1 2024.

Key objectives:
- Identify trending pain points
- Measure sentiment changes over time
- Recommend product improvements

When analyzing, prioritize recent feedback and cross-reference with NPS scores.

Related projects: Q4 2023 Analysis, Customer Satisfaction Study

​
Length Considerations

  • Concise: Aim for 100-500 words per memory
  • Focused: One topic or theme per memory
  • Scannable: Use line breaks for readability
  • Multiple Memories: Better to create several focused memories than one long one
Think of memories as index cards - each should contain a single coherent piece of information.

​
Structure

Use clear structure in your memory content: 
**Purpose**: What this project is about

**Context**: Background information

**Guidelines**: How to approach the content

**Important Notes**: Key things to remember

​
Clarity

  • Use simple, clear language
  • Avoid jargon unless necessary
  • Define terms that might be ambiguous
  • Be specific rather than vague

​
Editing Memories

1

Locate Memory

Find the memory in the Memories tab list.
2

Open Edit Dialog

Click the edit icon or “Edit” in the actions menu.
3

Modify Content

Update the name and/or content as needed.
4

Save Changes

Click “Save” to persist your changes.
Updated memory shows new content immediately.

​
Managing Memories

​
Delete Memory

Deleting a memory is permanent. The memory is removed from the project and AI models will no longer have access to it.
1

Select Memory

Click delete icon for the memory to remove.
2

Confirm Deletion

Review memory name and click “Delete” to confirm.
3

Verify Removal

Memory disappears from list immediately.

​
Memory Organization

​
Naming Strategy

Use consistent naming patterns: By Topic:
  • “Architecture - Overview”
  • “Architecture - Database Design”
  • “Architecture - API Structure”
By Type:
  • “[Guidelines] Code Style”
  • “[Context] Project History”
  • “[Reference] Related Projects”
By Priority:
  • “IMPORTANT: Budget Constraints”
  • “Key Finding: Customer Preference”
  • “Note: Technical Limitation”

​
Content Organization

Within memory content:
  • Start with most important information
  • Use consistent formatting
  • Include headers or sections
  • Add line breaks for readability

​
Next Steps

Best Practices

Write effective memories

MCP Servers

Provide AI access to your memories

Memories API

Manage memories programmatically