Using the Conversational Agent

This guide covers practical techniques for getting the best results from the VibeMap Conversational Agent. You will learn how to phrase requests, what kinds of prompts work well, and how changes are applied to your project.

Getting Started

Open any project in the editor and click the chat icon in the right panel to open the agent. The agent loads your full project context automatically -- you do not need to tell it which project you are working on.

Example Prompts

Adding Entities

Create new items by describing what you need:

• "Add a notification system feature that sends email and in-app alerts when a task is assigned"
• "Create a persona for an admin user who manages team permissions and billing"
• "Add three user stories for the search feature focused on filtering and sorting"
• "Generate acceptance criteria for the login feature covering email, Google, and GitHub sign-in"

Editing Entities

Modify existing items by referencing them directly:

• "Rename the Dashboard feature to Command Center"
• "Update the Developer persona to include API integration as a primary goal"
• "Change the priority of the Payment Processing feature to high"
• "Rewrite the acceptance criteria for user registration to include email verification"

Removing Entities

Delete items when they are no longer needed:

• "Remove the Guest User persona"
• "Delete the legacy export feature"
• "Remove all acceptance criteria from the Notifications feature so I can regenerate them"

The agent will show you an impact analysis before performing any deletion, so you can review what else might be affected.

Asking Questions

Query your project without making changes:

• "How many features does this project have?"
• "Which personas are not referenced by any user stories?"
• "What acceptance criteria exist for the authentication feature?"
• "Summarize the scope of this project in three sentences"

Bulk Operations

Handle multiple items in a single request:

• "Add acceptance criteria to every feature that is currently missing them"
• "Create user stories for all features tagged as MVP"
• "Update the priority of all billing-related features to high"

Tips for Getting Good Results

Be Specific

The more detail you provide, the better the output. Compare:

Reference Existing Entities

When editing or extending, name the entity you are referring to:

• "Add two more user stories to the File Upload feature"
• "Update the Power User persona to include mobile usage patterns"

Build Incrementally

You can chain requests across multiple messages. The agent remembers your conversation within the session:

1. "Add a reporting feature"
2. "Now add three user stories for it"
3. "Add acceptance criteria to each of those stories"

Ask for Confirmation When Unsure

If you are not sure how a change will play out, ask first:

• "What would happen if I removed the Notifications feature?"
• "Would adding a billing persona affect any existing user stories?"

How Changes Are Applied

When the agent performs an operation, changes are applied to your project in real time:

1. Immediate update -- The entity is created, modified, or removed in the database as soon as the agent confirms the operation.
2. Live UI refresh -- The project editor updates automatically. You will see new items appear, edits reflected, or removed items disappear without refreshing the page.
3. Conversation log -- Every change the agent makes is recorded in the chat history, so you have a full audit trail of what was modified and why.

Undo and Recovery

If the agent makes a change you did not intend:

• Ask the agent to reverse it -- "Undo the last change" or "Restore the persona you just deleted"
• Use regeneration -- Regenerate the affected section to get a fresh set of content
• Manual editing -- Click on any entity in the editor to edit it directly

Conversation Best Practices

• One topic per message -- Keep each message focused on a single request for the clearest results.
• Review before confirming -- When the agent asks for confirmation on destructive operations, read the impact analysis carefully.
• Start a new conversation if you switch to a completely different topic or the thread becomes long. This resets the context window and keeps responses focused.
• Use natural language -- You do not need special syntax or commands. Write as you would to a colleague.