Enable chatting with an Agentic Assistant agent from Slack.
Slack chats with the Agentic Assistant bridge the gap between where your team collaborates and where security operations happen by enabling you to interact with agents directly within you daily communication workflow without needing to log in to Cortex XDR.
Access the chat from Slack
Prerequisite
Before you can interact with the Cortex Agentic Assistant in Slack, the Slack v3 integration must be set up correctly. For more information, see the Slack v3 integration documentation.
To perform actions in Slack, your Slack email must match your Cortex XDR user email. This ensures the system can strictly follow your assigned permissions (RBAC). If you do not have the required permissions to interact with agents, the system will block the action.
Once the setup is done, you can initiate a chat or interact with the Cortex Agentic Assistant from Slack by tagging your configured bot name in a thread (for example @Your bot name). Cortex Agentic Assistant returns a dropdown menu of available agents to select.
Note
Only public agents are supported via Slack.
Considerations for Slack interactions with Agentic Assistant agents
When interacting with the Cortex Agentic Assistant directly within Slack, keep the following in mind to maintain session integrity and secure access:
Chat context
When an agent is tagged, the system automatically pulls in the last five messages in the thread (or up to the last bot interaction) so the agent understands the conversation's history.
Single player model
The first user to tag the bot becomes the initiator, and only this user can issue commands. If another user tries to send a prompt, they receive an Access Denied message.
The reset command
To hand off a session to another user or start fresh, any user in the Slack thread can type
@Your bot name reset. This ends the session and allows a new initiator to take over.Approving sensitive actions (hard locks)
Sensitive actions require approval. In Slack, this triggers a hard lock where the agent refuses text input and displays Approve (green) and Deny (red) buttons, and only the initiator can click these buttons. If the initiator does not respond within two weeks, the request is automatically denied and the chat will close.
Providing feedback
After a final result or remediation is executed, you can provide feedback directly in Slack using thumbs up or thumbs down.
Additional considerations
Small tables (less than 5 rows) are rendered as Markdown, while larger tables will be summarized with a link to view the full results. Code and logs will use standard Slack code blocks.
Chat artifacts are not visible via Slack.
Chat timeout
Slack sessions follow a timeout policy of two weeks of inactivity (matching the UI data retention policy), after which the session automatically closes.
Slack interaction with the Agentic Assistant example
The following is an example scenario describing how you can monitor shift priorities, track SLAs, and review pending automations in Cortex XDR directly from Slack.
Initiation
Check the daily queue by opening your team's Slack channel and tagging
@Your bot namewith the prompt, "What are the top issues I should prioritize today and show me all issues with an overdue SLA?".Agent selection
The bot responds with a dropdown menu of available public agents, and you select the appropriate agent to handle the request.
Status update
The agent processes the request and replies in the thread, providing a summarized list of the highest-priority issues and any automations currently waiting for user input.
Note
If a team member in the channel sees the summary and attempts to ask the agent, "Give me more details on the first SLA issue," the team member receives an access denied message because the active session is only available to you, the initiator.
Handoff
The session can remain open for up to two weeks, after which it automatically closes. To end a session, type
@Your bot name resetso the rest of the team can engage.Another team member can then tag
@Your bot nameto initiate a new session. Because the system pulls the last five messages in the thread, the agent understands the history of the conversation. The team member can simply prompt, "Assign the first overdue issue from that summary to me," and the agent will know which issue is being referenced.