Morpheus Adaptive Tasking

Last Updated: November 10, 2026

Overview

Morpheus Adaptive Tasking is the AI assistant in the incident workspace. It builds investigation plans from the incident, runs integration commands, summarizes incidents, and applies your team's standard operating procedures (SOPs) and rules. This article covers how to open it, its quick actions, the kinds of requests it handles, how it asks for approval before it acts, and common scenarios.

To open it, open an incident and click Open Morpheus (the Morpheus icon) at the bottom right of the incident workspace.

The Morpheus icon at the bottom right of the incident workspace

The Morpheus Adaptive Tasking panel opens beside the incident. Type a request in the chat input (Type / for quick actions, or ask anything…). The panel header has New session, Conversation history and Close sidebar buttons.

Administrators manage the rules, SOPs and action permissions that Morpheus follows in Configuration > Morpheus Guidance. See Morpheus Guidance.

Adaptive Tasking and Traditional Playbooks

Adaptive Tasking complements traditional pre-authored SOAR playbooks. The former generates investigation plans from the current incident context for novel or variable cases, where no matching playbook exists. The latter is well-suited for high-volume, regulated incident types where consistency matters most.

Quick Actions

Type / in the chat input to open the Quick Actions menu. Click a command, or keep typing to filter the list. To close the menu without running anything, press Esc. The menu closes and keeps what you typed, and opens again the next time you type /.

at-03-quick-actions.png

Group

Command

What it does

Work this incident

/Investigate

Build a step-by-step plan for this incident.


/Summarize

Brief someone on what is going on.


/RunPlaybook

Run an investigation playbook against this incident. Opens the playbook panel, where you pick the playbook to run.

Capture what you did

/Turn-into-SOP

Draft a reusable SOP from this conversation. The row shows how many exchanges there are to draft from.

Manage the library

/SOPs

Manage and view all Standard Operating Procedures. Personal and Shared below it list your personal SOPs and the shared SOPs.


/Rules

Rules Morpheus must always follow. Opens the Rule Management panel.


/Playbook

Browse saved Adaptive and Classic playbooks in Playbook Management.


/NewSession

Clear chat and start a new session.

/Investigate

When you run /Investigate, Morpheus:

  1. Reads the incident, including any earlier findings and runs on it.

  2. Searches your SOPs and uses one that fits the incident (see /SOPs).

  3. Builds a plan of up to eight steps. Each step names the integration and command that carries it out.

You can then run the plan's tasks, or ask Morpheus to regenerate the plan. A task that depends on another task waits until that task is done.

/Summarize

Generates a summary of the incident and the work done on it, including:

  • A headline and the verdict.

  • What happened, what was done and what is still pending.

  • An assessment and the key findings.

  • Recommended next steps.

A summary of the whole incident replaces the incident's AI summary, and it is saved as memory for the incident. Later investigations of similar incidents can find it. See Memory and Rules.

/RunPlaybook

Opens the playbook panel, where you pick an investigation playbook to run against the incident.

/Turn-into-SOP

Drafts an SOP from the current conversation. It needs at least one exchange with Morpheus on the incident; the menu row shows the count (for example, 1 exchange).

The draft opens in the New SOP window under Draft from this conversation. Click Choose what to capture to pick what goes into the draft. You can also open this window from + New SOP in SOP Management (see Method 2).

/SOPs

What is an SOP?

An SOP (standard operating procedure) is a written investigation procedure: the steps to take, and in which order, for a kind of incident.

Example phishing sop

  • Check sender domain reputation

  • Extract URLs and hashes

  • Query VirusTotal

  • Search SIEM for affected users

  • Check user interaction

  • Block sender if malicious

When Morpheus builds an investigation plan, it searches the active SOPs and follows one that fits the incident, combined with the incident's own data.

SOP Types

Scope

Created by

Visible to

Personal

Any user

Its creator

Shared

Administrators

All users (limited to the SOP's sites)

SOP Management

Choose /SOPs to open SOP Management. It lists the SOPs for the current incident type, grouped as PERSONAL and SHARED, with a search box (Search procedures...) and a status filter.

at-04-sop-management.png

How to Create SOPs

  • Method 1: Manual Creation

    1. Choose /SOPs, then click + New SOP.

    2. In the New SOP window, enter the procedure in SOP (Markdown; use Preview to check it). Then set:

      • SOP name (optional): if left empty, a name is made from the SOP.

      • Description (optional): when the SOP applies, in one sentence.

      • Category (optional).

      • Scope: Personal or Shared.

      • Incident Types: the incident types the SOP applies to.

      • Sites: the sites the SOP applies to.

    3. Click Create SOP.

  • Method 2: Draft from an Investigation

    1. Work the incident with Morpheus (at least one exchange).

    2. Choose /Turn-into-SOP, or click + New SOP and use Draft from this conversation > Choose what to capture.

    3. Review the draft, then click Create SOP.

at-05-new-sop-skill.png

How to Apply SOPs

Morpheus looks for a fitting SOP on its own when it builds a plan. To apply a specific SOP, click the SOP in SOP Management to open its details, then click Apply SOP.

An SOP's Details tab with the Active switch, Description, Content and Apply SOP

SOP Management Operations

Operation

Description

Edit (pencil)

Changes the SOP.

Enable / Disable (switch)

Turns the SOP on or off. A disabled SOP is not used when Morpheus looks for a fitting SOP.

Delete (trash)

Deletes the SOP. This cannot be undone.

History (in the details)

Lists the SOP's earlier versions.

/Rules

Opens Rule Management, where you add, view, turn on or off, and delete the rules Morpheus must always follow. See Memory and Rules.

/Playbook

Opens Playbook Management, which lists saved playbooks as cards on two tabs, Adaptive and Classic. Each card shows the playbook's name and its status with the last update (for example, Published · Updated Sep 22). A published playbook has a Run button on its card; the more-actions menu (...) holds the other actions. Draft playbooks show no Run button.

/NewSession

Clears the chat and starts a new session. Use it to start a new investigation or change direction. Morpheus asks you to confirm (Start new session). The session you leave stays under Conversation history in the panel header, where you can replay it (see Conversation History).

Outside the quick actions, you write requests in plain language. Morpheus reads each request and decides how to handle it: answer it, run a command, build or change a plan, summarize, search the web, or work with SOPs. These are the kinds of requests it handles, with example phrasings.

Type 1 - Ad-hoc Command Execution

Runs integration commands on request.

Example Use Cases

  • Checking IP, domain, or hash reputation

  • Searching for SIEM alerts

  • Blocking hosts

Example Phrasings

Goal

Example Phrasing

Checking IP reputation

"Check reputation for 8.8.8.8 using AbuseIPDB"

Searching SIEM alerts

"Search MS Sentinel for alerts related to lab1-pc1 in last 7 days"

Blocking a domain

"Block domain evil.com using Palo Alto"

Isolating an endpoint

"Quarantine endpoint DESKTOP-01 via CrowdStrike"

Performing a WHOIS lookup

"WHOIS lookup for suspicious-domain.com"

Executing multi-step operations

"First check IP 1.2.3.4 on VirusTotal, if malicious then block it on Fortinet"

Morpheus finds the command among your integrations and fills in its parameters. Read-only commands run straight away; commands that change something ask for approval first (see Action Approvals and Safety Guardrails). Naming the integration and the target helps it pick the right command.

Type 2 - Investigation Plan

Builds a structured investigation plan. Use /Investigate, or ask in plain language.

Example Phrasings

Goal

Example Phrasing

Generating a new plan

"Investigate this incident"

Regenerating a plan

"Regenerate the investigation plan"

Type 3 - Summary and Analysis

Analyzes the incident and the work done on it. /Summarize always produces the structured summary described in /Summarize.

Example Phrasings

Goal

Example Phrasing

Summarizing

"Summarize" or "Summarize this incident"

Analyzing

"Analyze the investigation results"

Type 4 - Web Search

Searches the internet for security information.

Example Use Cases

  • Looking up CVE details

  • Retrieving malware analysis reports

  • Retrieving attack technique descriptions

  • Accessing other relevant public information

Example Phrasings

Goal

Example Phrasing

Looking up CVE details

"Search the web for CVE-2024-1234"

Retrieving malware information

"Search for information about QakBot malware"

Retrieving attack technique details

"Look up MITRE T1059 on the web"

Say "on the web" when you want a web search. "Search Sentinel for alerts" is a request to run a command, not a web search.

Type 5 - SOP Management

Finds, applies or creates SOPs. Use /SOPs or /Turn-into-SOP, or ask in plain language.

Example Phrasings

Goal

Example Phrasing

Applying an SOP

Choose /SOPs, open an SOP, then click Apply SOP.

Creating an SOP from an investigation

"Create an SOP from this investigation"

Searching for SOPs

"Do we have an SOP for ransomware incidents?"

Type 6 - Task Execution

Runs tasks from the investigation plan.

The plan's tasks are listed on the TODO tab of the panel. You can run them one at a time with Run, or select several and run them together.

Running several tasks together

  1. Select the check box beside each task to run. A bar above the list shows how many are selected (for example, 3 selected), with Clear and Run 3.

  2. Click Run 3. The Run 3 tasks? window lists the tasks you selected.

  3. Click Run 3 tasks.

The TODO tab with three tasks selected and the 3 selected bar with Clear and Run 3
r2-4-confirm-run.png

The selected tasks run without asking again. As the window warns, any of them that changes something will do so without a further approval prompt. Click a task to review its full details before you select it.

The selected tasks run together. When they finish, one Task Run card shows the results grouped by task. Its summary has one line per task, giving the task's verdict (for example, benign, malicious or inconclusive) and how many commands it ran. Below that, each task has its own section with what its command returned. The tasks then move to the DONE tab.

r2-4-grouped-results.png
r2-4-grouped-results-tasks.png

A task whose check box is turned off cannot be selected yet. Hover over the box for the reason. For example, a task that needs other tasks' results shows Blocked until its prerequisites finish.

If Morpheus needs more information to run a task, for example which connection to use, it asks in the chat. Pick an answer to continue.

Example Phrasings

Goal

Example Phrasing

Running the next task

"Run the next task"

Continuing execution

"Continue" or "Go ahead"

Type 7 - Security Knowledge Q&A

Answers questions on security topics without running anything.

Example Use Cases

  • Asking about security concepts

  • Asking about attack techniques

  • Asking about SOC best practices

  • Asking about how to use the D3 platform

Example Phrasings

Goal

Example Phrasing

Accessing ATT&CK knowledge

"What is MITRE T1059?"

Exploring ATT&CK techniques

"How does PsExec lateral movement work?"

Reviewing best practices

"What's the best practice for investigating phishing?"

Understanding platform usage

"What integrations do we have available?"

Morpheus can also draw on memory, such as summaries of similar incidents, and on your SOPs. See Memory and Rules.

Action Approvals and Safety Guardrails

What happens when Morpheus wants to run a command depends on what the command does and on the site's action-permission policy:

  • Read-only commands (searches, lookups, reputation checks) run without asking.

  • Commands that change something (for example, adding a tag, blocking, quarantining or closing) stop at an approval card. The card names the command and lists its parameters. Click Approve to run it, or Reject to cancel it.

  • A command the policy blocks does not run.

After the command runs, a Command · [command name] card shows the result: the command, its status (for example COMPLETED), what it did, and View Details and Evidence.

at-02-assistant-panel.png

In this example the request "add a new tag named 'testbyysl'" needed approval: Approve running D3 internal.AddD3IncidentTags? This action requires approval under the site's action-permission policy. After Approve, the Command · [command name] card shows the command completed.

Result Cards

A result card is the box Morpheus adds to the chat each time it finishes a piece of work. Each card's title names what the work produced, so you can scan a long conversation and find the summary, the plan or a specific command without opening every card. The status badge (RUNNING, COMPLETED) stays at the right of the title.

Result Card Titles

Card title

What produced it

Incident Summary

/Summarize, or asking for a summary

Investigation Plan

/Investigate, or asking for a plan

SOP Plan: [SOP name]

Applying an SOP (Apply SOP in the SOP's details)

Task Run · [number] tasks

Running tasks from the plan

Command · [command name]

A command Morpheus ran at your request

While a card is still running, its title shows the noun alone (Task Run, SOP Plan). The detail that is only known at the end, such as the number of tasks or the SOP's name, is added when the card completes.

Task Steps

While a task runs, its card shows the steps Morpheus is taking. Three common steps are described in plain words:

Step

What Morpheus is doing

Checking sources

Reading the reference material it needs for the task

Working through tasks

Listing the plan's tasks to pick up the next one

Saving results

Updating a task's status with its result

Other steps keep their existing names.

Conversation History

Morpheus keeps every session you have on an incident. Click Conversation history in the panel header to list your earlier sessions on this incident, grouped by day. Each row shows the session's first request as its title, when it was last active, and how many messages and actions it holds. Use Search conversations and actions… to find a session. The list shows only your own sessions.

The Conversation History list with one earlier session, its title, time, message count and action count

Replaying a session

Click a session to replay it. The replay is read-only: a banner reads Read-only replay, and you continue the work in a new session. Two views sit at the top:

  • Conversation: the session's messages and result cards, as they appeared.

  • Actions: each command Morpheus ran in the session, one per row, with the integration, its class (for example, Read or Write) and the time. The footer totals the session's actions, approvals and blocked attempts. An action that needed approval also shows who approved it and when, for example Approved by you followed by the date and time.

The Actions view listing four read-only commands with their integration and time
An approved Write action in the Actions view, marked Approved by you with the approval time, and the footer counting 1 approval

Click an action to see its details: the integration, when it was requested and run, and the command's Parameters and Output. Click the back arrow to return to the list.

r2-12-action-detail.png

Common Scenario Walkthroughs

Scenario 1 - New Incident

User initiates a full incident investigation. The system generates a multi-step plan, executes tasks, and produces a summary report, with an option to save the workflow as an SOP.

  1. Enter /Investigate in the chat.

  2. Review the generated plan.

  3. Run the plan's tasks.

  4. Enter /Summarize in the chat to generate a report.

  5. (Optional) Choose /Turn-into-SOP to save the investigation as an SOP.

Scenario 2 - Quick IOC Check

User requests an IOC reputation check. The system finds the command, runs it, and returns results.

Example flow

User: Check reputation for hash 44d88612fea8a8f36de82e1278abb02f using VirusTotal.

Adaptive Tasking: Runs the read-only command → Returns VT analysis results

Scenario 3 - SIEM Alert Search

User requests alert search in MS Sentinel. The system runs the query and returns matching alerts.

User: Search MS Sentinel for alerts related to 192.168.1.100 in the last 24 hours

Adaptive Tasking: Runs the read-only search → Returns alert list

Scenario 4 - Security Knowledge Question

User asks about a MITRE ATT&CK technique. The system provides a direct answer without executing actions.

Example flow

User: What is MITRE ATT&CK technique T1059?

Adaptive Tasking: Provides a direct explanation without executing any actions

Scenario 5 - D3 Platform Operation

User requests an incident update. The system identifies the operation, asks for approval, and applies the change.

Example flow

User: Change the severity of this incident to Critical

Adaptive Tasking: Identifies as D3 incident operation → Approval card (Approve) → Updates severity

Scenario 6 - New Incident Type Without SOP

An ad-hoc investigation plan generated from incident context.

Example

Situation: A PowerShell execution alert occurs on host FIN-WS-04 by user john.doe, with no matching SOP.

Execution:

  • User enters /Investigate.

  • The system finds no fitting SOP and generates a plan using available integrations and context.

  • Example plan:

    1. Enrich the user via Active Directory.

    2. Retrieve the PowerShell command from EDR.

    3. Extract IOCs from the command.

    4. Check IOC reputation.

    5. Search related SIEM activity.

TAKEAWAYS

  • No SOP required

  • Steps align with available integrations

  • Intermediate data processing is inserted automatically

  • Contextual fields are pre-populated

After execution, the investigation can be saved as an SOP with /Turn-into-SOP.

Scenario 7 - Novel Incident Without a Playbook

An ad-hoc investigation plan generated for an unfamiliar technique.

Example

Situation: An alert mapped to MITRE T1528 (Steal Application Access Token) with no existing SOP or playbook.

Execution:

  • User enters /Investigate.

  • The system builds a plan using MITRE mapping and available incident data (user, application, scopes, source IP).

  • Example plan:

    1. Retrieve OAuth app grants for the target user (last 30 days).

    2. Analyze granted scopes and identify high-privilege permissions.

    3. Enrich the source IP using a reputation service.

    4. Search for related sign-in activity from the IP across all users.

    5. Review mailbox rules and forwarding settings for persistence.

TAKEAWAYS

  • No existing SOP or playbook required

  • Steps align with available integrations

  • Data transformation steps are added automatically

  • Relevant context fields are auto-filled