Loading...
Loading...
Guides, templates, and tools for developers
Universal Cursor IDE setup with AI subagents for development automation. Includes 10 specialized subagents, 5 command workflows, task management system, and coding standards.
This video is available after purchase
Please sign in to purchase
By clicking the button, you agree to the terms of the Public Offer and Privacy Policy
AI-assisted development often means chaos: subagents work in the background, context gets lost, you don't know what's happening. This configuration solves the problem — all subagents work sequentially in one chat, you see every step, full control.
Cursor AI Setup — ready-to-use Cursor IDE configuration with an AI subagent system for development automation. Just download, copy to your project, and get a complete task orchestration system with automatic checks, tests, and documentation.
🔒 One-time purchase — lifetime access
Pay once, no subscriptions. Download and use in all your projects forever.
🔄 Free updates
All new versions, additional subagents, and improvements are free after purchase.
💬 Feedback and requests
Have ideas for additions? Found a bug? Contact me:
I constantly improve the setup based on user feedback.
| Component | What it is | Why you need it |
|---|---|---|
| 10 AI Subagents | Specialized subagents for different tasks | Worker creates code, Test-Runner tests, Debugger fixes, Documenter documents |
| 5 Command Workflows | /implement, /orchestrate, /refactor, /review, /audit | From simple tasks to full features with planning, tests, and review |
| 11 Skills | Reusable modules for subagents | Orchestration, task management, git-helper, documentation |
| 5 Rules | Universal coding standards | Commit, testing, documentation, security standards |
| Hook System | Event-based automation (optional) | Auto-format, lint-fix, run tests after changes |
| Command | When to use | Workflow |
|---|---|---|
/implement | Simple tasks (component, function) | Code → Test → Docs |
/orchestrate | Complex tasks (features, systems) | Plan → Loop[Code → Test → Review] → Docs |
/refactor | Code works but is poorly organized | Analyze → Refactor → Verify → Docs |
/review | Before commit or PR | Review → (Fix) → Verify |
/audit | Full project health check | Architecture → Security → Quality → Report → (Fix) |
All commands automatically invoke the right subagents in the correct order.
/implementWhen you need to quickly create a component, function, or endpoint:
/implement Create a React Button component with label and onClick props
Happens automatically:
Result: ready component with tests and documentation in minutes.
/orchestrateWhen you need a full feature with planning and automatic checks:
/orchestrate Add authentication system with email/password and OAuth
Happens automatically:
Phase 1: Planning
Phase 2: Execution — for each task:
Phase 3: Finalization
Result: complete feature with automatic fixes, checks, and documentation.
/refactorWhen code works but is poorly organized:
/refactor Improve structure of src/services/payment-processor.ts
Happens automatically:
Important: tests must exist before refactoring. If not — use /implement to add tests first.
/review/review src/components/Form/
By default checks staged changes (git diff --staged). For non-critical findings — asks whether to fix.
/audit/audit
Happens automatically:
When to use: before a major release, onboarding into a new codebase, periodic health check.
All subagents can be used both automatically (via workflows) and manually for specific tasks.
| Subagent | Command | What it does | Used in |
|---|---|---|---|
| Worker | /worker | Creates code | /implement, /orchestrate |
| Planner | /planner | Breaks tasks into subtasks | /orchestrate |
| Test Runner | /test-runner | Tests + verification | /implement, /orchestrate, /refactor, /review |
| Test Writer | /test-writer | Writes tests for existing code | manually |
| Debugger | /debugger | Fixes errors and bugs | all workflows |
| Reviewer | /reviewer | Code quality review | /orchestrate, /review, /audit |
| Documenter | /documenter | Creates documentation | /implement, /orchestrate, /refactor, /audit |
| Refactor | /refactor | Improves code structure | /refactor, /audit |
| Security Auditor | /security-auditor | Security audit | /orchestrate, /audit |
| Senior Reviewer | /senior-reviewer | Architectural review | /refactor, /audit |
Subagents use skills to perform complex tasks:
| Skill | Purpose |
|---|---|
| simple-workflow | Basic workflow for /implement |
| orchestration | Full orchestration for /orchestrate |
| refactor-workflow | Safe refactoring workflow |
| review-workflow | Code review workflow |
| audit-workflow | Full health check workflow |
| task-management | Task and plan management |
| docs | Documentation generation |
| git-helper | Git operations and commits |
| security-guidelines | Security checks |
| architecture-principles | Architectural patterns |
| code-quality-standards | Code quality standards |
Rules apply to all subagents:
| Rule | What it covers |
|---|---|
| commit-messages | Commit format (Conventional Commits) |
| git-workflow | Git branching and PR workflow |
| testing | Test and coverage requirements |
| documentation | Code documentation standards |
| security | Basic security requirements |
Workspace (Temporary) — .cursor/workspace/
Documentation (Permanent) — configurable paths
No duplication, parallel orchestrations without conflicts, recovery after failure.
Edit .cursor/config.json:
{
"documentation": {
"paths": {
"root": "docs",
"plans": "docs/plans",
"reports": "docs/reports",
"issues": "docs/issues"
},
"enabled": {
"plans": true,
"reports": true,
"issues": false
}
}
}
Using docs/ instead of ai_docs/? Change "root": "docs". Don't need reports? Set "reports": false.
configure-agents.md contains ready-to-use prompts for auto-configuring subagents. Copy the prompt into Cursor chat, and the subagent will scan your project and update the files:
.cursor/workspace/
.cursor/
├── agents/ # 10 specialized subagents
│ ├── worker.md # Code creation
│ ├── planner.md # Task planning
│ ├── test-runner.md # Testing
│ ├── test-writer.md # Writing tests
│ ├── debugger.md # Error fixing
│ ├── reviewer.md # Code review
│ ├── documenter.md # Documentation
│ ├── refactor.md # Refactoring
│ ├── security-auditor.md # Security audit
│ └── senior-reviewer.md # Architectural review
│
├── commands/ # 5 automatic workflows
│ ├── implement.md # /implement
│ ├── orchestrate.md # /orchestrate
│ ├── refactor.md # /refactor
│ ├── review.md # /review
│ └── audit.md # /audit
│
├── skills/ # 11 reusable modules
│ ├── simple-workflow/
│ ├── orchestration/
│ ├── refactor-workflow/
│ ├── review-workflow/
│ ├── audit-workflow/
│ ├── task-management/
│ ├── docs/
│ ├── git-helper/
│ ├── security-guidelines/
│ ├── architecture-principles/
│ └── code-quality-standards/
│
├── rules/ # 5 universal standards
│ ├── commit-messages.md
│ ├── git-workflow.md
│ ├── testing.md
│ ├── documentation.md
│ └── security.md
│
├── config.json # Configuration (adapt to your project!)
├── hooks.json # Hooks (optional)
└── workspace/ # Temporary data (don't commit)
/implement Create Card component in src/components/Card.tsx
Result: component + tests + documentation
/orchestrate Add Stripe payment system
Result: task plan → automatic implementation → tests → fixes → code review → documentation
/orchestrate @TODO.md
Planner uses your file, updates task statuses (⏳→🔄→✅) directly in it.
/refactor Improve structure of src/services/payment-processor.ts
/test-runner # Verify nothing broke
/review src/components/Form/
/security-auditor Check auth endpoints in src/api/auth/
/worker Fix bug in validateEmail in utils/validation.ts
/test-runner
/senior-review Evaluate caching design in @CACHE-DESIGN.md
# Safe refactoring manually
/refactor src/payment/processor.ts
/test-runner
/security-auditor src/payment/
Optionally configure hooks for automation:
Details in HOOKS.md with AI prompts for creating hooks.
Q: Which command should I use?
A: Writing code — /implement (simple) or /orchestrate (complex). Improving existing code — /refactor. Checking before commit — /review. Want the full picture — /audit.
Q: Can I use subagents manually?
A: Yes! All subagents can be called directly: /worker, /debugger, /refactor, /security-auditor, etc. Workflows just automate chains of calls.
Q: Do subagents run in the background?
A: No, everything runs sequentially in one chat. You see every step.
Q: Can I disable documentation?
A: Yes, in .cursor/config.json set the relevant flags to false or paths to null.
Q: Can I combine subagents my own way?
A: Yes! Call subagents in any sequence. For example: /refactor → /test-runner → /security-auditor.
Q: What if orchestration was interrupted?
A: Everything is saved in .cursor/workspace/failed/, resumable in a future update.
Q: I don't see the .cursor folder in the downloaded archive
A: Folders and files starting with a dot are hidden by default. Enable hidden files display:
Cmd + Shift + . in Finderdocs/, ai_docs/, your own structureThis is a universal configuration. Files do NOT contain:
This is a foundation that needs to be supplemented with instructions for your project.
After download:
.cursor/ folder to project root.cursor/config.json – configure documentation pathsconfigure-agents.md.cursor/workspace/ to .gitignore/implement, /orchestrate, /refactor, /review, /audit⚠️ The
.cursor/folder is hidden. On macOS:Cmd + Shift + .in Finder. On Windows: Explorer → View → Show → Hidden items. In VS Code/Cursor the files are visible in the sidebar.
Visibility – all subagents work in one chat, you see every step
Control – you can stop, adjust, or continue manually
Flexibility – use workflows for automation or subagents directly
Universality – works with any stack and documentation structure
Extensibility – add your own subagents, skills, rules
Download the configuration, copy to your project, configure paths in config.json and subagents via configure-agents.md – and start automating development from the first command.