The MCP Servers provides insights into MCP servers accessed by your organization steered by Netskope.
Source – Agentic Broker and Netskope Client
- Category – Shows the top 10 MCP server categories by usage. Click a category to filter.
- CCL – Shows the distribution of servers by Cloud Confidence Level. Click a bar to filter.
The following table displays a list of MCP Servers discovered in your environment with the following details:
- Name – Name of the MCP server.
- Footprint – Number of managed endpoints where this MCP server was detected by Netskope Client. Hover over or click on the footprint number to see more details. Available when Endpoint AI Discovery is enabled. For details on what Netskope Client discovers on the endpoint, see Assets Detected during AI Discovery.
- MCP Category – Category of the MCP server. For example, AI Infrastructure, Communication, Database.
- CCL Score – Cloud Confidence Level associated with the server.
- Risk – Reconciled risk level for the application, combined with a confidence indicator. Hover over the confidence dots to see the confidence level. This data comes from the AI Risk Agent (Beta), part of AI SecOps. For details on how risk is calculated, see the AI Risk Agent documentation in your tenant. See What Does the Risk Score and Confidence Level Mean in AICC?
- #d Identities – Number of unique users interacting with the MCP server over the selected time range.
- #d Events – Number of events triggered on the server for selected time range.
- #d Sessions – Number of sessions for the selected time range.
- First Seen – Timestamp when the server was first detected.

MCP Server Detail Panel
The MCP server detail panel shows what AI Command Center (AICC) knows about a single MCP server: its metadata, how users, devices, agents, and models connect to it, the findings and risk evaluation for it, where it runs, the policies it triggered, and the identities that use it. To open the panel, click any row in the MCP servers table. The panel opens on the right side of the page. Click the close icon (X) at the top right of the panel to close it.

Panel Header
The header stays visible at the top of the panel as you scroll. It contains the following details:
- MCP server name – Name of the MCP server, with an icon.
- Type – Type of the item. Always MCP server.
- First seen – Date when the server was first detected.
- Last seen – Date of the most recent detection.
Metadata
The Metadata card shows the identifying details of the MCP server:
- MCP category – Category of the server.
- Server type – Type of the server.
- Protocol version – Version of the Model Context Protocol (MCP) that the server uses.
- CCL – Cloud Confidence Level score of the server, shown with a cloud icon.
Viewing Server Connections
The Connection card shows a diagram of the users, devices, and assets that connect to the MCP server. The selected server appears highlighted in the diagram. Each node shows a label below it with a sample name and the number of additional items. A label on a line shows the data volume between two nodes.
Click the expand icon at the top right of the Connection card to open the connection graph in a larger view. The view shows the number of findings with a severity bar, followed by one card for each finding.

Viewing Findings
The Findings card lists the security findings that AICC raised for the MCP server. The number next to the card title is the total number of findings. A bar below the title shows the split by severity, with a legend that lists the count for each severity, for example, Medium 3, Low 2, and Informational 1.
The card shows up to five findings. Each row shows a severity dot, the finding title, and the time since the finding was last evaluated, for example, 2h ago. Examples of findings include:
- Unusually heavy or long MCP usage
- MCP used on a device with an unsupported operating system
- MCP server allowed to use the client’s AI model
- MCP server only partly covered by policy
- MCP connection using an outdated protocol version
- User copied many prompt templates from an MCP server
Some rows show a Posted label with a ticket icon instead of the time.
View all (n) – Click to see the full list of findings.
Footprint
The Footprint card shows whether the MCP server is a local install on a managed endpoint or a remote server consumed over the network. The label next to the card title shows the time range.
- Endpoint: Servers that run locally on a managed endpoint. They are declared in a local AI-client configuration with a local command or a loopback (localhost) address. For more information, see Device Classification.
- Cloud: Servers that users access remotely. Netskope detects them in proxy traffic or in a local AI-client configuration that uses a hosted URL.
Click a row to open its list in the same panel.
Viewing Risk
The Risk card shows the risk evaluation for the MCP server. The subtitle reads Risk evaluation sourced from AISecOps. For details on how AICC calculates risk, see What Does the Risk Score and Confidence Level Mean in AICC?
- Risk level – Risk level of the server, for example, High or Medium. This is the same level as in the Risk column of the table.
- Evaluation time – Date and time of the evaluation, for example, Nov 4, 1 PM.
- Risk summary – A text summary of why the server has this risk level, followed by key counts, for example, the number of DLP incidents and how many were blocked at policy.
- Create case – Click to create an AI risk case for the server in AISecOps. While AICC creates the case, the button shows Creating case. After the case exists, the button changes to the case number and status. For details about cases, see AISecOps documentation in your tenant.
- Previously created cases – Expand the n previously created case row to see earlier cases for this server. Each row shows the creation date, the case title, the case status, for example, Investigated, and the case number. Click a row to open the case.
Viewing Policy Violations, Sessions, and Identities
Policy violations – The Policy violations card lists the Netskope policies that the server’s traffic triggered. The number next to the card title is the number of violations, and the label shows the time range. Each row shows the policy name, the policy type tag, and the number of events. The card shows up to five policies, ordered by number of events, highest first. Click View all (n) to see the full list. When there are no violations, the card shows the message No policy violations.
Sessions – The Sessions card shows a chart of the sessions that the MCP server handled over the selected time range. The total appears next to the card title. Hover over the chart to see the number of sessions at a specific date and time.
Identities – The Identities card lists the identities that use the MCP server. The number next to the card title is the total number of identities. Each row shows an identity, a bar that compares the identities, and a value. The card shows up to five identities, with the highest value first. Click View all (n) to see the full list.
Viewing Prompts, Tools, Resources, and Connected Clients
The bottom of the panel shows what the server offers and who connects to it.
- Prompts – Prompts that the server provides. The number next to the title is the total. Displays No prompts available when AICC finds none.
- Tools – Number of tools that the server exposes.
- Resources – Number of resources that the server exposes.
- Connected clients – Number of MCP clients that connect to the server.

