Feature request for product/service
Cursor IDE
Describe the request
Problem Statement
Currently, when Cursor Agent successfully completes a complex task (like implementing authentication, setting up a CI/CD pipeline, or building a specific feature), there’s no way to capture that workflow and reuse it in other projects. Each new project starts from scratch, and we have to re-explain the same process, architectural decisions, and step-by-step approaches.
Proposed Solution
Introduce a Workflow Memory System that allows Cursor Agent to:
Record Workflows - Automatically or manually capture the sequence of actions, decisions, and steps the agent took to complete a task
Save as Templates - Store these workflows with descriptive names (e.g., “Add JWT Authentication”, “Set up Next.js with Prisma”, “Implement Dark Mode”)
Replay in New Projects - Apply saved workflows to different projects with contextual adaptation
Use Case Example
Project A: Agent successfully implements a complete authentication system with:
JWT token management
Protected routes
User registration/login flows
Password hashing with bcrypt
Refresh token rotation
Email verification
Project B (New): Instead of re-explaining all this, I could simply say:
“Use the authentication workflow from Project A”
And the agent would:
Adapt the same architectural approach
Follow the same security best practices
Implement similar patterns adjusted for the new tech stack
Apply the same testing strategy
How It Could Work:
Option 1: Manual Workflow Saving
User: “Save this conversation as a workflow called ‘E-commerce Checkout Flow’”
Agent: “Workflow saved! You can apply it in future projects with @E-commerce-Checkout-Flow”
Option 2: Automatic Workflow Detection
Agent detects a successful multi-step feature implementation and suggests:
“I’ve completed the payment integration. Would you like to save this as a reusable workflow?”
Option 3: Workflow Library
A dedicated section in Cursor Settings showing:
My Workflows - Personal saved workflows
Team Workflows - Shared across organization (Team/Enterprise plans)
Community Workflows - Public workflows shared by the community
Benefits:
Consistency - Same best practices across all projects
Time Savings - No need to re-explain complex processes
Knowledge Capture - Preserve successful patterns that worked
Team Collaboration - Share proven workflows across teams
Onboarding - New team members can leverage existing workflows
Continuous Improvement - Iterate and refine workflows over time
Difference from Current Rules Feature
While Rules allow us to define instructions and preferences, they require:
Manual documentation of the workflow
Generic instructions that may not capture the full context
No connection to actual successful implementations
Workflow Memory would:
Capture the actual execution pattern from real implementations
Include context about decisions made during the process
Be derived from successful completions, not written beforehand
Store the “how” not just the “what”
Integration with Existing Features
This could work alongside:
Rules - Workflows could auto-generate project rules
Hooks - Workflows could trigger specific hooks
Chat History - Link to original conversations for reference
Chat Export - Export workflows as documentation
Technical Considerations
Workflows should be stored with metadata (tech stack, dependencies, date created)
Smart adaptation needed when tech stacks differ between projects
Privacy controls for sensitive workflows
Version control for workflow evolution
Import/export functionality for sharing
Questions for the community:
Would you use this feature? What workflows would you save first?
Should workflows be sharable publicly (like a workflow marketplace)?
How should the agent handle conflicts when a workflow doesn’t perfectly fit the new project context?
Should there be a limit on workflow length/complexity?
This feature would transform Cursor Agent from a helpful coding assistant into a true knowledge accumulation system that gets smarter with every project you complete. Instead of ephemeral conversations, we’d build a library of battle-tested workflows that compound our productivity over time.