Usage Guide

Learn how to use all of Routa’s features.

Table of contents

  1. Opening Routa
  2. Multi-Agent Sessions
    1. Creating Sessions
    2. Managing Sessions
  3. Chat Interface
    1. Input Area
    2. Keyboard Shortcuts
  4. Message Types
  5. Working with Tools
    1. Tool Call Visualization
    2. Permission Requests
  6. Example Workflows
    1. Code Review
    2. Refactoring
    3. Writing Tests
  7. Tips and Best Practices

Opening Routa

  1. Open the Routa tool window
    • Find it in the right sidebar
    • Or: View → Tool Windows → Routa
  2. The Welcome page shows:
    • Detected agents
    • Input area for chatting
    • Agent selector dropdown

Multi-Agent Sessions

Creating Sessions

  • Connect to different agents simultaneously - each gets its own tab
  • Switch between agents using the toolbar or tabs
  • Each session maintains independent conversation history

Managing Sessions

  • New Session: Click + Add Agent or select a different agent
  • Switch Session: Click on tabs or use the agent selector
  • Close Session: Click the × on the tab or use Disconnect

Chat Interface

Input Area

The input area at the bottom provides:

  • Text Area: Type your message here
  • Agent Selector (bottom-left): Switch agents on the fly
  • Send Button (bottom-right): Or press Enter to send
  • Stop Button: Cancel ongoing requests

Keyboard Shortcuts

Shortcut Action
Enter Send message
Shift+Enter New line in message
Esc Stop current request

Message Types

Routa displays different message types with distinct styling:

💬 User Messages (Blue)
Your messages to the agent
🤖 Assistant Messages (Gray)
Agent responses and explanations
⚡ Tool Calls (Orange)
Agent executing tools (file operations, terminal commands, etc.)
✅ Tool Results (Green/Red)
Results from tool executions (green for success, red for errors)
💭 Thinking (Purple)
Agent’s internal reasoning process
ℹ️ Info Messages (Cyan)
System messages and notifications
⚠️ Error Messages (Red)
Error messages and warnings

Working with Tools

Tool Call Visualization

When an agent uses tools, you’ll see:

  1. Tool Call - What the agent is doing 
    [Tool Call] Reading file: src/main/java/UserService.java
  2. Tool Result - The outcome 
    [Tool Result] ✓ File content retrieved successfully

Permission Requests

Some tools require permission:

  1. Agent requests permission to perform an action
  2. A dialog appears asking for approval
  3. Choose Allow or Deny

Set autoApprove: true in config to automatically approve all requests for trusted agents.

Example Workflows

Code Review 

You: Can you review the changes in UserController.java?

Agent: I'll review the file for you.
[Tool Call] Reading file: src/main/java/UserController.java
[Tool Result] ✓ File read successfully

Agent: Here's my review:
1. The error handling looks good...
2. Consider adding validation for...

Refactoring 

You: Help me refactor the authentication logic

Agent: I'll analyze the current implementation first.
[Tool Call] Reading file: src/auth/AuthService.java
[Tool Result] ✓ File read successfully

Agent: I suggest extracting the token validation into a separate method...

Writing Tests 

You: Write unit tests for the PaymentService class

Agent: I'll create comprehensive tests for PaymentService.
[Tool Call] Reading file: src/service/PaymentService.java
[Tool Result] ✓ File read successfully
[Tool Call] Creating file: src/test/PaymentServiceTest.java
[Tool Result] ✓ Test file created successfully

Agent: I've created tests covering all payment scenarios...

Tips and Best Practices

Be Specific: Provide clear context and specific requests for better results.

Use Multiple Agents: Different agents have different strengths - try them for different tasks.

Review Tool Calls: Always review what tools the agent is using, especially for file modifications.

Save Important Conversations: Copy important responses before closing sessions.