CCI Insights Agent - Netskope Technical Documentation

The article introduces the CCI Insights Agent, a Gen AI tool designed to streamline application risk assessment for administrators.

You can enable the agent on a tenant from October 12, 2026.

Current Challenges in Risk Assessment

Administrators currently rely on time-consuming and inefficient methods to gather information about applications.
These methods include performing web searches, manually searching App Catalogs, navigating through filters, or contacting support teams.
With more than 86,000 apps in the catalog, even one question becomes a multi-step research project.

CCI Insights Agent Overview

The agent empowers administrators to ask natural-language queries regarding individual applications or the broader catalog. It retrieves data directly from the App Catalog’s data lake and, when necessary, consults vendor-specific sources, ensuring seamless risk assessment without requiring the user to leave the tenant environment.
It is located within the App Catalog section of the dashboard, accessible through an icon in the bottom right corner.
The tool provides AI-powered insights and enables natural language search to get relevant insights in no time.

How the Agent Functions

Natural Language Queries: Users can enter prompts, such as requesting a list of applications within a specific category (for example, “Gen AI”) or across the whole catalog. The agent retrieves the data and presents it in a descriptive summary accompanied by a table.
Contextual Follow-ups: The agent retains the context of previous questions, allowing users to ask follow-up questions to further refine results without needing to restate the initial parameters.
Feedback Loop
- If the agent is unable to answer a query, users can provide feedback to help refine and improve its performance for future use cases.
- The interface includes controls to view chat history, allowing users to continue previous sessions or initiate new ones.

By utilizing this tool, administrators can significantly reduce the time and effort required to perform risk assessments on application portfolios.

Sample Questions

Single-app lookups: “Does Box support SSO?”, “Is instance detection supported for App X?”
Catalog-wide, risk-based cuts: “Give me a list of applications in the GenAI category,” “Which apps in the Finance category have a High or Excellent CCL but don’t encrypt data at rest, and have any of those vendors publicly announced plans to add it?”
Conversational refinement of a prior answer: “From this, give me a list of Google applications with more than 80 CCL score”
Compliance: reviewing third-party apps’ risk posture ahead of an audit deadline

Current Challenges in Risk Assessment

Administrators currently rely on time-consuming and inefficient methods to gather information about applications.

These methods include performing web searches, manually searching App Catalogs, navigating through filters, or contacting support teams.

With more than 86,000 apps in the catalog, even one question becomes a multi-step research project.

CCI Insights Agent Overview

The agent empowers administrators to ask natural-language queries regarding individual applications or the broader catalog. It retrieves data directly from the App Catalog’s data lake and, when necessary, consults vendor-specific sources, ensuring seamless risk assessment without requiring the user to leave the tenant environment.

It is located within the App Catalog section of the dashboard, accessible through an icon in the bottom right corner.

The tool provides AI-powered insights and enables natural language search to get relevant insights in no time.

How the Agent Functions

Natural Language Queries: Users can enter prompts, such as requesting a list of applications within a specific category (for example, “Gen AI”) or across the whole catalog. The agent retrieves the data and presents it in a descriptive summary accompanied by a table.

Contextual Follow-ups: The agent retains the context of previous questions, allowing users to ask follow-up questions to further refine results without needing to restate the initial parameters.

Feedback Loop

If the agent is unable to answer a query, users can provide feedback to help refine and improve its performance for future use cases.
The interface includes controls to view chat history, allowing users to continue previous sessions or initiate new ones.

By utilizing this tool, administrators can significantly reduce the time and effort required to perform risk assessments on application portfolios.

Sample Questions

Single-app lookups: “Does Box support SSO?”, “Is instance detection supported for App X?”

Catalog-wide, risk-based cuts: “Give me a list of applications in the GenAI category,” “Which apps in the Finance category have a High or Excellent CCL but don’t encrypt data at rest, and have any of those vendors publicly announced plans to add it?”

Conversational refinement of a prior answer: “From this, give me a list of Google applications with more than 80 CCL score”

Compliance: reviewing third-party apps’ risk posture ahead of an audit deadline