---
title: "Working with engagements"
description: "How engagements organize related DrowAI work and preserve durable context."
section: "Workflow"
updated: 2026-06-18
tags: ["engagements","workflow"]
source: "/user-guide/working-with-engagements"
---

# Working with engagements

How engagements organize related DrowAI work and preserve durable context.

Engagements are the top-level workspace for a body of authorized security work.
They keep related tasks, knowledge, evidence, findings, and reports connected
to the same assessment context.

## How tasks join engagements

Every DrowAI task belongs to an engagement. Engagements are not an optional
extra layer; they are the container DrowAI uses to keep task activity,
knowledge, evidence, findings, and reports connected.

If a user creates a task without first creating or selecting an engagement,
DrowAI automatically creates an engagement for that task using the task's name
and description. The task then becomes part of that engagement.

Create or select an engagement deliberately when multiple task sessions should
contribute to the same assessment or report. For example, one engagement may
contain separate tasks for discovery, service review, web assessment, and final
validation.

## What belongs in an engagement

- A clear engagement name.
- A short description of the assessment context.
- One or more scoped tasks.
- Durable knowledge collected from task activity.
- Engagement reports generated from selected task inputs.

## Working pattern

1. Decide whether the task should join an existing engagement or start a new engagement.
2. Create a task with clear scope. If no engagement is selected, DrowAI creates one automatically from the task.
3. Start each task when you are ready to run work.
4. Use chat, workspace files, shell access, and knowledge views to review task output.
5. Stop tasks when useful runtime work is complete.
6. Prepare selected tasks as report inputs.
7. Generate and review the engagement report draft.

## Archiving

Archive an engagement when active runtime work is finished but preserved
knowledge should remain available. DrowAI blocks archive while runtime-active
tasks still exist, so stop or retire active tasks first. Archiving is a
lifecycle action for the workspace, not a signal to discard evidence or report
history. Archived engagements can be restored when work needs to continue.