Skip to content

memory-bank-guide.md

Latest commit

 

History

History
215 lines (160 loc) · 5.52 KB

memory-bank-guide.md

File metadata and controls

215 lines (160 loc) · 5.52 KB

CursorRIPER

CursorRIPER Framework - Memory Bank Guide

The Memory Bank is a core component of the CursorRIPER Framework, providing persistent context across coding sessions. This guide explains how to effectively use and maintain your Memory Bank.

Memory Bank Overview

flowchart TD
    A[Memory Bank] --> B[projectbrief.md]
    A --> C[systemPatterns.md]
    A --> D[techContext.md]
    A --> E[activeContext.md]
    A --> F[progress.md]
    
    B -->|Project<br>Foundation| G[AI Assistant]
    C -->|Architecture<br>& Patterns| G
    D -->|Technical<br>Context| G
    E -->|Current<br>Status| G
    F -->|Progress<br>Tracking| G
    
    G -->|Updates| E
    G -->|Updates| F
Loading

Core Memory Files

projectbrief.md

Purpose: Defines core requirements and goals of the project.

Key Sections:

  • Project Overview
  • Core Requirements
  • Success Criteria
  • Scope (In/Out)
  • Timeline
  • Stakeholders

When to Update:

  • At the beginning of the project
  • When major requirements change
  • When project scope evolves

systemPatterns.md

Purpose: Documents system architecture and design patterns.

Key Sections:

  • Architecture Overview
  • Key Components
  • Design Patterns in Use
  • Data Flow
  • Key Technical Decisions
  • Component Relationships

When to Update:

  • After architectural decisions
  • When new patterns are established
  • When component relationships change

techContext.md

Purpose: Describes technologies used and development setup.

Key Sections:

  • Technology Stack
  • Development Environment Setup
  • Dependencies
  • Technical Constraints
  • Build and Deployment
  • Testing Approach

When to Update:

  • When adding new technologies
  • After updating dependencies
  • When changing build processes
  • When modifying development environment

activeContext.md

Purpose: Tracks current work focus and immediate next steps.

Key Sections:

  • Current Focus
  • Recent Changes
  • Active Decisions
  • Next Steps
  • Current Challenges
  • Implementation Progress

When to Update:

  • At the start of each work session
  • After completing significant tasks
  • When switching focus areas
  • When encountering new challenges

progress.md

Purpose: Tracks what works, what's in progress, and what's left to build.

Key Sections:

  • Project Status
  • What Works
  • What's In Progress
  • What's Left To Build
  • Known Issues
  • Milestones

When to Update:

  • After completing features
  • When starting new work
  • When discovering issues
  • When reaching milestones

Memory Bank Management

Automatic Updates

The framework can automatically update memory files based on your customization settings:

  • AUTO_UPDATE_MEMORY: Controls whether automatic updates are enabled
  • MEMORY_UPDATE_FREQUENCY: Controls when updates occur

Manual Updates

You can manually update memory files by:

  1. Editing the files directly
  2. Asking the AI to update specific files
  3. Using the memory bank update command

Example:

Please update the progress.md file to reflect that we've completed the user authentication feature.

Memory Bank Best Practices

  1. Keep Information Current

    • Regularly update activeContext.md and progress.md
    • Ensure technical decisions are documented in systemPatterns.md
    • Update project requirements in projectbrief.md as they evolve

  2. Use Consistent Formatting

    • Follow the templates for each file type
    • Use markdown formatting consistently
    • Include dates for all updates

  3. Be Specific in Updates

    • Include specific file paths, function names, and component names
    • Document issues with detailed descriptions
    • Link related items across files

  4. Maintain Version History

    • Include version numbers in memory files
    • Update the "Last Updated" timestamp when making changes
    • Track major revisions

  5. Organize Information Logically

    • Use clear section headers
    • Group related information together
    • Use lists for items that belong together

Memory Bank and RIPER Workflow

Each RIPER mode interacts with the Memory Bank in specific ways:

RESEARCH Mode

  • Reads from: projectbrief.md, systemPatterns.md, techContext.md
  • Updates: activeContext.md with new discoveries

INNOVATE Mode

  • Reads from: all memory files
  • Updates: systemPatterns.md with design alternatives and decisions

PLAN Mode

  • Reads from: all memory files
  • Updates: activeContext.md with planned changes, progress.md with expected outcomes

EXECUTE Mode

  • Reads from: all memory files
  • Updates: activeContext.md with progress, progress.md with completed items

REVIEW Mode

  • Reads from: all memory files
  • Updates: progress.md with review findings, activeContext.md with review status

Troubleshooting Guide Link

If you encounter issues with your Memory Bank:

  1. Missing Information

    • Check if files exist in the memory-bank directory
    • Verify file permissions
    • Re-run memory bank initialization if necessary

  2. Outdated Information

    • Update timestamp and version number
    • Review and update content in all relevant sections
    • Consider a full memory bank refresh

  3. Inconsistent Information

    • Cross-reference across memory files
    • Update all related files when making changes
    • Ask the AI to help reconcile inconsistencies

  4. File Corruption

    • Restore from backups
    • Regenerate files from templates if necessary
    • Ask the AI to help recreate missing information

The CursorRIPER Framework prevents coding disasters while maintaining perfect continuity across sessions.