Skip to content

Commands

Jump to bottom
iDorgham edited this page Nov 23, 2025 · 2 revisions

Commands Reference

Complete reference for all DoPlan CLI commands. Commands are used within your IDE to interact with the AI agent system.

Command Overview

DoPlan CLI uses a slash-command system (/command) that activates specific AI agents to help with your project. Commands are organized into three categories:

  • Core Commands - Essential commands for the development workflow
  • Team Commands - Commands for managing and viewing agents
  • Specialized Commands - Domain-specific commands for advanced use cases

Core Commands

/tell - Capture Project Idea

Description: Capture your project idea and save it to the project plan.

Usage:

/tell
/tell <idea>

Examples:

/tell
/tell Build a todo app with authentication and dark mode
/tell Create a REST API for a blog platform

What it does:

  1. Captures your project idea (inline or prompts for it)
  2. Saves to .plan/00_System/IDEA.md
  3. Activates Project Orchestrator and Product Manager
  4. Updates project state

Files Modified:

  • .plan/00_System/IDEA.md
  • .plan/active_state.json

Agents Involved:

  • Project Orchestrator
  • Product Manager

Next Steps: After capturing your idea, use /improve to brainstorm.

/improve - Team Brainstorm Session

Description: Get ideas and improvements from all Level 1 managers.

Usage:

/improve

Examples:

/improve

What it does:

  1. Activates all Level 1 managers (Product, Engineering, Design, QA, Release, Documentation)
  2. Conducts a brainstorm session
  3. Saves all ideas to .plan/00_System/BRAINSTORM.md

Files Read:

  • .plan/00_System/IDEA.md

Files Modified:

  • .plan/00_System/BRAINSTORM.md

Agents Involved:

  • Product Manager
  • Engineering Lead
  • Design & UX Manager
  • QA & Reliability Manager
  • Release & Growth Manager
  • Documentation Lead

Next Steps: Review BRAINSTORM.md, then use /write to generate planning documents.

/write - Generate Planning Documents

Description: Generate comprehensive planning documents (PRD, Architecture, Design System).

Usage:

/write

Examples:

/write

What it does:

  1. Product Manager creates PRD.md
  2. Engineering Lead and System Architect create ARCHITECTURE.md
  3. Design Manager and UI/UX Designer create DESIGN_SYSTEM.md
  4. Saves all documents to .plan/00_System/

Files Read:

  • .plan/00_System/IDEA.md
  • .plan/00_System/BRAINSTORM.md

Files Modified:

  • .plan/00_System/PRD.md
  • .plan/00_System/ARCHITECTURE.md
  • .plan/00_System/DESIGN_SYSTEM.md
  • .plan/active_state.json

Agents Involved:

  • Product Manager
  • Engineering Lead
  • System Architect
  • Design & UX Manager
  • UI/UX Designer
  • Project Orchestrator

Next Steps: Review the documents, use /change to edit, or /good to approve.

/change - Edit Documents

Description: Edit any planning document with natural language instructions.

Usage:

/change <document> <change>

Examples:

/change prd Add dark mode support
/change architecture Use PostgreSQL instead of MySQL
/change design Add mobile-first responsive design

What it does:

  1. Parses the document name and change description
  2. Loads the specified document
  3. Applies the requested changes
  4. Saves the updated document

Files Read:

  • .plan/00_System/*.md

Files Modified:

  • .plan/00_System/*.md (specified document)

Agents Involved:

  • Project Orchestrator

Document Names:

  • prd - Product Requirements Document
  • architecture - Technical Architecture
  • design - Design System

/good - Approve & Lock Plan

Description: Approve the planning documents and lock them for task generation.

Usage:

/good

Examples:

/good

What it does:

  1. Validates that PRD.md, ARCHITECTURE.md, and DESIGN_SYSTEM.md exist
  2. Sets locked: true in active_state.json
  3. Updates phase to "approved"

Files Read:

  • .plan/00_System/PRD.md
  • .plan/00_System/ARCHITECTURE.md
  • .plan/00_System/DESIGN_SYSTEM.md
  • .plan/active_state.json

Files Modified:

  • .plan/active_state.json

Agents Involved:

  • Project Orchestrator

Next Steps: Use /tasks to generate implementation tasks.

/tasks - Generate Implementation Tasks

Description: Generate organized implementation tasks from the approved plan.

Usage:

/tasks

Examples:

/tasks

What it does:

  1. Reads the approved plan (PRD, Architecture, Design System)
  2. Generates implementation tasks organized by phases
  3. Creates .plan/TASKS.md
  4. Updates phase to "tasks"

Files Read:

  • .plan/00_System/PRD.md
  • .plan/00_System/ARCHITECTURE.md
  • .plan/00_System/DESIGN_SYSTEM.md
  • .plan/active_state.json

Files Modified:

  • .plan/TASKS.md
  • .plan/active_state.json

Agents Involved:

  • Project Orchestrator
  • Engineering Lead
  • Product Manager

Next Steps: Review TASKS.md, then use /build to start coding.

/build - Start Coding

Description: Start implementing the next task or a specific task.

Usage:

/build
/build <task_id>

Examples:

/build
/build 1.2
/build 3

What it does:

  1. Determines the task (next uncompleted or specific task_id)
  2. Loads task context, dependencies, and related code
  3. Activates relevant agents (Frontend Lead, Backend Lead, etc.)
  4. Begins implementation
  5. Updates active_task in active_state.json

Files Read:

  • .plan/TASKS.md
  • .plan/active_state.json

Files Modified:

  • .plan/active_state.json
  • Source code files

Agents Involved:

  • Engineering Lead
  • Project Orchestrator
  • Relevant team leads (based on task)

GitHub Automation: After task completion, changes are auto-committed and pushed.

Next Steps: Complete the task, then use /finished to mark it complete.

/progress - Track Progress

Description: Show current project progress and status.

Usage:

/progress

Examples:

/progress

What it does:

  1. Reads TASKS.md and active_state.json
  2. Calculates progress metrics:
    • Total tasks
    • Completed tasks
    • In progress tasks
    • Percentage complete
  3. Displays formatted progress report

Files Read:

  • .plan/TASKS.md
  • .plan/active_state.json

Agents Involved:

  • Project Orchestrator

Output Example:

Phase: building
Tasks: 15/30 completed (50%)
Current task: 1.2 - Create API endpoints
Next up: 1.3 - Implement authentication

/finished - Complete Task

Description: Mark the current task as complete and auto-commit changes.

Usage:

/finished

Examples:

/finished

What it does:

  1. Marks task complete in TASKS.md
  2. Updates active_state.json (removes active_task, adds to completed)
  3. Auto-commits changes with conventional commit format
  4. Auto-pushes to current branch
  5. Updates CHANGELOG.md if significant changes

Files Read:

  • .plan/TASKS.md
  • .plan/active_state.json

Files Modified:

  • .plan/TASKS.md
  • .plan/active_state.json
  • CHANGELOG.md (if applicable)

Agents Involved:

  • Project Orchestrator
  • Release Captain

GitHub Automation:

  • Auto-commit with conventional commit format
  • Auto-push to current branch (feature/bugfix/hotfix)
  • Update CHANGELOG.md
  • Follow branching strategy

Next Steps: Use /build to start the next task.

Team Commands

/team - Show Agents

Description: Display the hierarchical structure of all AI agents and their roles.

Usage:

/team

Examples:

/team

What it does:

  1. Loads all agent definitions from .cursor/agents/
  2. Displays the hierarchical structure
  3. Shows each agent's role and responsibilities

Files Read:

  • .cursor/agents/*.md

Agents Involved: None (display only)

Output: Shows the complete agent hierarchy with roles.

/load - Inject Context

Description: Load rules or files into agent context for the current session.

Usage:

/load <path>

Examples:

/load @library/04-frameworks/nextjs.md
/load .plan/00_System/PRD.md
/load @library/07-databases/postgresql.md

What it does:

  1. Parses the file or directory path
  2. Loads the content
  3. Injects it into agent context for the current session

Files Read:

  • .cursor/rules/library/**
  • .plan/**

Agents Involved:

  • Project Orchestrator

Use Cases:

  • Load framework-specific rules
  • Load database rules
  • Load planning documents for reference
  • Load custom rules

Specialized Commands

/ship - Release Management

Description: Manage releases, versioning, and deployment.

Usage:

/ship

Examples:

/ship

What it does:

  • Coordinates release process
  • Manages versioning
  • Handles deployment workflows

Agents Involved:

  • Release Captain
  • Release & Growth Manager

/safe - Security Audit

Description: Perform security audit and review.

Usage:

/safe

Examples:

/safe

What it does:

  • Reviews code for security issues
  • Checks for vulnerabilities
  • Provides security recommendations

Agents Involved:

  • Security Lead

/cheap - Cost Optimization

Description: Optimize costs and resource usage.

Usage:

/cheap

Examples:

/cheap

What it does:

  • Analyzes resource usage
  • Identifies cost optimization opportunities
  • Provides recommendations

Agents Involved:

  • DevOps Engineer
  • Performance Engineer

Command Tips & Tricks

1. Workflow Order

Follow this order for best results:

  1. /tell - Capture idea
  2. /improve - Brainstorm
  3. /write - Generate plans
  4. /good - Approve
  5. /tasks - Generate tasks
  6. /build - Start coding
  7. /finished - Complete tasks

2. Using /change Effectively

Be specific with your change requests:

  • ✅ Good: /change prd Add user authentication with OAuth2
  • ❌ Bad: /change prd fix it

3. Building Specific Tasks

You can build any task by ID:

/build 1.2    # Build task 1.2
/build 3.5     # Build task 3.5

4. Loading Context

Load relevant rules before building:

/load @library/04-frameworks/nextjs.md
/build

5. Tracking Progress

Check progress regularly:

/progress

6. Combining Commands

You can use multiple commands in sequence:

/tell Build a blog
/improve
/write
/good
/tasks
/build

Command Syntax

General Syntax

/command [arguments]

Arguments

  • Required arguments: Must be provided
  • Optional arguments: Can be omitted
  • Inline arguments: Can be provided inline or prompted

Examples

# No arguments
/tell
/improve
/write

# With inline arguments
/tell Build a todo app
/change prd Add dark mode

# With task ID
/build 1.2

Command Reference Table

Command Category Arguments Auto-commit
/tell Core Optional: idea No
/improve Core None No
/write Core None No
/change Core Required: doc, change No
/good Core None No
/tasks Core None No
/build Core Optional: task_id No
/progress Core None No
/finished Core None Yes
/team Team None No
/load Team Required: path No
/ship Specialized None No
/safe Specialized None No
/cheap Specialized None No

Related Pages

Last Updated: 2025
Maintained By: Documentation Team

Clone this wiki locally