Explain View (Graph-Level AI Assistant)

Overview

The Explain View feature in CAST Imaging Express helps users understand complex architectural graph views by transforming them into clear, business-friendly explanations. It is designed to support architects, developers, managers, and stakeholders who need rapid insight into application structure without deep technical analysis.

How to access

Open any graph view in CAST Imaging Express.
Click the Explain button at the bottom-right of the view.
Read the popup — it confirms that only data from the currently visible graph is sent to the AI. You can dismiss it permanently using Don’t show if desired.
Interact with the chatbot window that opens at the bottom-right. It presents an initial explanation and a prompt box for follow-up questions.

Use the download button in the chatbot header to review the exact data sent to the AI for the current view.

Purpose & capabilities

Explain View provides:

Business-friendly summaries — High-level explanations of what the current view represents and why it matters.
Relationship explanations — Descriptions of how components, layers, and nodes connect and interact.
Interactive Q&A — Follow-up questions scoped strictly to the currently visible view.

This feature is intended to accelerate understanding, support communication, and improve knowledge transfer.

Typical AI response structure

Each AI-generated explanation includes:

Section	Description
Purpose	A brief statement describing what the view represents.
Overview	A high-level explanation of how major components relate.
Component Breakdown	The roles of key node types and their relationships.
Top Insights	A short list of prioritized takeaways for stakeholders.

What data is used

Transmitted to the AI

Graph metadata (node names, types, and relationships)
Layout coordinates (to support spatial and contextual descriptions)
Only the currently visible subset of the graph view

Never transmitted

Source code content
Proprietary business logic or detailed implementations
Credentials, secrets, or sensitive configuration values

No source code is ever sent to the AI. Only structural metadata from the visible graph view is processed.

Processing flow

User View → Graph Metadata Extraction → Claude Sonnet → Business-Friendly Explanation

Best practices

Zoom or filter the view to the most relevant scope before clicking Explain to get targeted results.
Use follow-up questions to clarify specific relationships or components.
Download and review the transmitted payload when validating data boundaries with stakeholders.

Limitations

Context is limited to the visible view; switching views resets the context.
Explanations reflect metadata only — they do not infer hidden business rules from source code.
AI does not validate architectural correctness.

Recommended usage scenarios

Scenario	How the AI Helps
Executive briefings	Concise, non-technical summaries for stakeholders.
Architecture reviews	Accelerates comprehension of relationships and dependencies.
Knowledge transfer	Self-service exploration and consistent explanations across views.
Pre-migration analysis	Business-level application overview.

By using Explain View, you acknowledge that structured graph metadata may be transmitted to an AI service. You are responsible for ensuring that the visible graph scope is appropriate for AI-assisted analysis.

These features are governed by CAST’s Artificial Intelligence and Machine Learning Terms.