The CursorRIPER Framework is designed to be highly customizable to match your preferences and project needs. This guide explains how to customize the framework's behavior through the customization.mdc file.
flowchart TD
A[customization.mdc] --> B[Response Style]
A --> C[Mode Behavior]
A --> D[Memory Management]
A --> E[Archive Behavior]
A --> F[Command Aliases]
A --> G[Mode Extensions]
A --> H[Documentation Preferences]
A --> I[Custom Project Structure]
The customization file is located at:
.cursor/rules/customization.mdc
This file is loaded after core framework components and can override default behaviors.
Control how the AI assistant communicates with you:
RESPONSE_VERBOSITY: "BALANCED"
Possible values:
"CONCISE": Short, direct responses with minimal explanation"BALANCED": Moderate detail with clear explanations (default)"DETAILED": Comprehensive responses with extensive context
CODE_STYLE_PREFERENCES: ""
Specify your coding style preferences, such as:
- Indentation (tabs vs. spaces, size)
- Naming conventions (camelCase, snake_case, etc.)
- Line length limitations
- Bracket placement
- Comment style
EXPLANATION_LEVEL: "MEDIUM"
Possible values:
"MINIMAL": Basic explanation of code functionality"MEDIUM": Balanced explanation of code with key implementation details (default)"COMPREHENSIVE": Detailed explanation of all aspects of the code
Customize how modes operate and transition:
SUGGEST_MODE_TRANSITIONS: true
If true, the AI can suggest when a mode transition might be appropriate.
AUTO_MODE_TRANSITION: false
If true, the AI can automatically transition between modes (except to EXECUTE).
EXECUTE mode always requires explicit user authorization.
PLAN_QUESTION_COUNT: 5
Number of clarifying questions the AI will ask in PLAN mode before drafting a plan.
Control how and when memory files are updated:
AUTO_UPDATE_MEMORY: true
If true, the AI will automatically update memory files after significant changes.
MEMORY_UPDATE_FREQUENCY: "AFTER_COMPLETION"
Possible values:
"AFTER_EVERY_RESPONSE": Update after each AI response"AFTER_COMPLETION": Update after completing a significant task (default)"MANUAL_ONLY": Only update when explicitly requested
REQUIRED_MEMORY_FILES: ["projectbrief.md", "activeContext.md", "progress.md"]
List of memory files that must exist for the framework to function properly.
Configure how archiving works:
AUTO_ARCHIVE_START_PHASE: true
If true, the START phase will be automatically archived upon completion.
BACKUP_FREQUENCY: "DAILY"
Possible values:
"NEVER": Don't create automatic backups"DAILY": Create backups daily (default)"WEEKLY": Create backups weekly"BEFORE_CHANGES": Create backups before making significant changes
KEEP_BACKUP_COUNT: 5
Number of backup sets to retain before deleting the oldest.
Create shortcuts for mode transition commands:
CUSTOM_COMMANDS: {
"/r": "/research",
"/i": "/innovate",
"/p": "/plan",
"/e": "/execute",
"/rev": "/review"
}
Define your own aliases for frequently used commands.
Add custom behaviors to each RIPER mode:
RESEARCH_MODE_EXTENSIONS: []
INNOVATE_MODE_EXTENSIONS: []
PLAN_MODE_EXTENSIONS: []
EXECUTE_MODE_EXTENSIONS: []
REVIEW_MODE_EXTENSIONS: []
Each array can contain strings that define additional behaviors for the respective mode.
Add custom phases and workflows:
CUSTOM_PHASES: []
Additional project phases beyond the standard ones.
CUSTOM_WORKFLOWS: []
Custom workflows for specific project types.
Configure how documentation is generated:
DOCUMENTATION_STYLE: "MARKDOWN"
Format for generated documentation.
INCLUDE_CODE_COMMENTS: true
Whether to include detailed comments in generated code.
CODE_BLOCK_LANGUAGE_TAGS: true
Whether to include language tags in code blocks.
Configure how the AI formats its responses:
MODE_DECLARATION_FORMAT: "[MODE: {mode}]"
Format string for mode declarations at the start of responses.
PROGRESS_INDICATOR_FORMAT: "[{current_step}/{total_steps}]"
Format for progress indicators in responses.
Define project-specific settings:
PROJECT_TYPE: "DEFAULT"
Identifies the type of project for specialized handling.
CUSTOM_FOLDER_STRUCTURE: {}
Custom folder structure definitions for project scaffolding.
TECHNOLOGY_PRESETS: {}
Predefined technology stacks for quick selection.
Here are examples of common customization scenarios:
RESPONSE_VERBOSITY: "CONCISE"
EXPLANATION_LEVEL: "MINIMAL"
This configuration will make the AI provide shorter, more direct responses with minimal explanations.
SUGGEST_MODE_TRANSITIONS: true
AUTO_MODE_TRANSITION: true
This allows the AI to suggest and automatically perform mode transitions (except to EXECUTE mode, which always requires explicit authorization).
CUSTOM_COMMANDS: {
"/rs": "/research",
"/in": "/innovate",
"/pl": "/plan",
"/ex": "/execute",
"/rv": "/review",
"/s": "/start"
}
These shortcuts provide quicker access to mode transitions.
PROJECT_TYPE: "FRONTEND"
TECHNOLOGY_PRESETS: {
"FRONTEND": {
"languages": ["JavaScript", "TypeScript", "HTML", "CSS"],
"frameworks": ["React", "Vue", "Angular"],
"buildTools": ["Webpack", "Vite"]
}
}
Specialized configuration for frontend projects.
To apply changes to your customization settings:
- Edit
.cursor/rules/customization.mdcwith your preferred settings - Save the file
- Start a new chat session in Cursor IDE
The AI assistant will load your custom settings and acknowledge the active customizations in its first response of each session.
To restore default settings:
- Delete the custom values from
.cursor/rules/customization.mdc - Replace with default values or leave blank to use framework defaults
- Start a new chat session in Cursor IDE
For advanced use cases, consider:
- Custom Mode Extensions: Create specialized behaviors for specific project types
- Custom Phases: Design new phases for specialized development workflows
- Integration with CI/CD: Configure the framework to work with your CI/CD pipeline
- Custom Memory Files: Add project-specific memory files for specialized information
- Start Simple: Begin with minimal customization and add more as needed
- Document Changes: Note why you've customized specific settings
- Version Control: Include customization file in version control for team consistency
- Regular Review: Periodically review customizations to ensure they still meet your needs
- Team Alignment: Ensure team members understand and follow customized workflows
The CursorRIPER Framework prevents coding disasters while maintaining perfect continuity across sessions.
