Module 14 • Advanced

Multi-Agent Workflows

Discover how JiXe coordinates multiple AI agents (Kodi, Mark, Rei, Tif, Trey, Dex) to accomplish complex projects in parallel.

📊 Module Progress

Complete objectives and quizzes to unlock your badge!
0/9 Tasks Completed

🎯 Learning Objectives

Identify Agent Roles — Memorize the 7 agents (JiXe, Kodi, Mark, Rei, Tif, Trey, Dex) and their specialties.
When to Use Multi-Agent — Decide between single-agent vs multi-agent workflows.
Describe the Workflow Steps — The 4-step process: breakdown, spawn, collect, review.
Use sessions_spawn — Understand how JiXe spawns sub-agents for isolated work.

📚 Theory

The Agent Team (7 Total)

JiXe isn't alone — there's a whole team of AI agents ready to work for you!

Agent Role Specialty
JiXe 💻 Head Coordinator Overall management, delegation
Kodi 🧩 Coding Specialist Writing, debugging, refactoring code
Mark 📣 Marketing Specialist Content creation, campaigns
Rei 🔍 Research Specialist Deep research, data gathering
Tif 🤳 Social Media Specialist Social content, scheduling
Trey 📈 Trading Specialist Market analysis, trading strategies
Dex 📊 Data Analyst Data processing, visualization

When to Use Multiple Agents

Scenario 1: Complex Project
You: "Build me a new website for AI course, research competitors, write content, and deploy it."
JiXe's plan: Rei → research, Mark → write content, Kodi → code website, Tif → social posts, JiXe → coordinate.

Scenario 2: Parallel Tasks
You: "Research 3 AI tools, summarize this PDF, and fix the bug in my code — all within 1 hour."
JiXe spawns: Rei (research), Dex (summarize), Kodi (fix bug). All run in parallel — saves time!

📚 How Multi-Agent Workflows Work

Step-by-Step Process

Step 1: You give the task to JiXe.
You: "Research AI tools, then write a comparison blog post."

Step 2: JiXe breaks it down: Task 1 (Research → Rei), Task 2 (Write → Mark).

Step 3: JiXe spawns sub-agents. They work in isolated sessions and return results.

Step 4: JiXe reviews, combines, and delivers the final result to you.

Technical Implementation

JiXe uses the sessions_spawn tool to create sub-agents:

{
  "task": "Research top 5 AI tools",
  "runtime": "subagent",
  "label": "research-task"
}

The sub-agent works in an isolated session (doesn't clutter your main chat).

Key Takeaway

Multi-agent workflows = divide and conquer. JiXe coordinates, specialists execute. Use it for complex, multi-step projects!

🧪 Hands-On Exercise

Design a multi-agent workflow for a real project.

Step 1: Choose Your Project

Pick one: Launch a new product • Analyze a dataset • Build an app • Custom: __________

Step 2: Break Down Tasks

List the steps needed:
1. __________
2. __________
3. __________
4. __________

Step 3: Assign to Agents

Match each task to the best agent:
Task 1 → Agent: ____ (Reason: ____)
Task 2 → Agent: ____ (Reason: ____)
Task 3 → Agent: ____ (Reason: ____)
Task 4 → Agent: ____ (Reason: ____)

Step 4: Consider Dependencies

Which tasks depend on others? (e.g., Task 2 depends on Task 1)

Step 5: Write the Workflow

Example:
Step 1: JiXe spawns Rei to research competitor products
Step 2: JiXe spawns Mark to write product description (uses Rei's research)
Step 3: JiXe spawns Kodi to build product landing page
Step 4: JiXe spawns Tif to create social media campaign
Step 5: JiXe reviews all and deploys

📝 Quiz: Multi-Agent Workflows

Q1: Who coordinates all other agents?

A
Kodi
B
JiXe
C
Mark
D
Rei

Q2: Sub-agents work in:

A
The main chat session
B
Isolated sessions
C
Another computer
D
A parallel universe

Q3: Which agent is best for coding tasks?

A
Mark
B
Kodi
C
Tif
D
Trey

Q4: What tool does JiXe use to spawn sub-agents?

A
create_agent
B
sessions_spawn
C
magic_summon
D
please_start

Q5: Multi-agent workflows are best for:

A
Simple questions
B
Complex, multi-step projects
C
Playing games
D
Wasting time
← Previous
Module 13
Hub
🤖 AI Hub
Complete Module to Unlock 🔒
Module 15
🤝
Badge Earned!
Multi-Agent Ready 🤝

You've completed all objectives and quizzes!