Claude Code Prompts

Important

These prompts are designed to work effectively but please use them with caution and review any code you ship to production

Prompts

• Multi-agent
• More to come :)

How to use

1. Take a prompt you like and copy it like Multi-agent
2. Run the command nano ./claude/commands/prompt-name.md (For this example you would put multi-agent.md)
3. Paste in the prompt
4. Click Ctrl + X then y and finally enter
5. Go to a directory where you want to run claude code
6. Run claude code through the claude command
7. Once you see it load in type /prompt-name [YOUR TASK HERE] (For the multi-agent example you would type /multi-agent and then your task
8. Once you typed task you can click enter and watch claude code work!

Warning

By default it will commit code it writes once it's done so if you want to disable this just add to your prompt Do not commit any code I will do this myself if needed

Breakdown of prompts

Multi-agent

The multi-agent prompt transforms Claude into an Autonomous Orchestration Agent that breaks down complex coding tasks into manageable sub-tasks and executes them in parallel.

Visual Workflow: View Multi-agent Flow Diagram

Core Process:

1. Task Analysis - Analyzes your codebase and the complex task you've given it
2. Smart Decomposition - Breaks the large task into 3-7 independent sub-tasks that can run in parallel
3. Creates Task Folder - Generates a structured folder with naming like Feature - Task Name - 2025-07-07
4. Generates Individual Plans - Creates detailed markdown files for each sub-task with:

• Mission briefs and implementation guidelines
• Todo checklists for analysis, design, implementation, testing, and documentation
• Risk assessments and testing strategies

5. Master Coordination - Creates an info.md file with:

• Overall strategy and detected technical guidelines
• Git protocols for coordination between sub-agents
• Dependencies and collaboration rules

6. Parallel Execution - Launches ALL independent sub-tasks simultaneously as a batch operation
7. Centralized Git Management - Reviews and commits each sub-task's work separately with proper commit messages

Key Features:

• True Parallelism: Independent tasks run simultaneously, not sequentially
• Dependency Management: Handles task dependencies intelligently
• Git Protocol: Centralized commit management prevents conflicts
• Adaptive Technology Detection: Automatically detects your project's tech stack and follows existing patterns
• Quality Control: Each sub-task includes comprehensive testing and review steps

This approach is perfect for large features, refactoring projects, or any complex development task that can benefit from parallel execution and systematic coordination.