| name | do |
| description | Quick actions for common tasks |
{{#if args}}
🧠 AI Router - Auto Agent Selection
Task: {{args}}
Step 1: Analyze Task
First, analyze the task and determine which agents/commands to use.
Available Agents:
Development
| Agent | Command | Use When |
|---|
| Planner | /plan | Need implementation plan |
| Scout | /scout | Find files, understand codebase |
| Coder | /code | Write/modify code |
| Debugger | /debug | Fix bugs, errors |
| Tester | /test | Generate/run tests |
Quality
| Agent | Command | Use When |
|---|
| Reviewer | /review | Code review, security check |
| PR Review | /review-pr | Review pull request |
DevOps
| Agent | Command | Use When |
|---|
| Git Manager | /git | Git operations |
| PR Creator | /pr | Create pull request |
| DB Admin | /db | Database operations |
Documentation
| Agent | Command | Use When |
|---|
| Docs Manager | /docs | Update documentation |
| Copywriter | /copywrite | Marketing content |
Research/Creative
| Agent | Command | Use When |
|---|
| Researcher | /research | Learn new topics |
| Brainstormer | /brainstorm | Generate ideas |
| Designer | /design | UI/UX design |
Step 2: Route Decision
Based on the task "{{args}}", you need to:
- Identify task type (bug fix? new feature? docs? research?)
- Select 1-4 agents in execution order
- Explain your routing decision
- Execute the first agent, then ask if user wants to continue with next agent
Common Patterns:
- New feature: Planner → Scout → Coder → Tester → Reviewer
- Bug fix: Debugger → Tester
- Refactor: Scout → Coder → Reviewer
- Documentation: Docs Manager
- Research: Researcher → Brainstormer → Planner
Step 3: Execute
Now analyze the specific task "{{args}}" and:
- State which agents you will use and why
- Start executing as the FIRST agent in the sequence
- At the end, tell user about the next recommended agent
Begin analysis and execution:
{{else}}
🧠 AI Router - Auto Agent Selection
Automatically analyze your task and run the right agents.
Usage:
/do <describe your task>
Examples:
/do Create a login form with validation
/do Fix the null pointer error in user.ts
/do Write tests for the payment module
/do Refactor the API to use async/await
/do Document the authentication flow
How It Works:
- AI analyzes your task
- Selects 1-4 appropriate agents
- Executes them in optimal order
- Coordinates handoffs between agents
Agent Categories:
- Development: Planner, Scout, Coder, Debugger, Tester
- Quality: Reviewer, PR Review
- DevOps: Git, PR, Database
- Docs: Documentation, Copywriting
- Research: Brainstorm, Research, Design
Just describe what you want to do!
{{/if}}