How to use Vaiz AI: prompts, tasks and documents


Vaiz AI helps you turn project information into tasks, reports, and documents. Start with a ready-made prompt or describe what you need, add the relevant context, and review the result.

Use the Vaiz AI assistant to ask about project work and prepare drafts your team can review. Steps 1-5 explain how to write and refine a request. Steps 6-8 show ways to continue, using separate examples of task creation, project documents, and an optional external AI connection.

Step 1. Open the chat and choose a starting point

Open the AI chat in Vaiz. The start screen offers ready-made prompts for common requests:

Prompt
What it helps you do
My tasks
View the tasks assigned to you.
Deadlines
Find overdue tasks and work due this week.
High priority
Save a Smart List of your top-priority tasks.
Weekly report
Prepare a board summary with charts.
Team standup
Review recent updates grouped by team member.
Milestone progress
Check progress toward your milestones.
Team workload
View a chart of open tasks by assignee.
Catch me up
Catch up on notifications and recent activity.
Create a task
Draft a new task with the assistant.
The My tasks card is highlighted as one starting point.

Select a card that fits your goal. To start with a custom request, type in the Ask anything... field.

You can add details as the conversation develops.

Step 2. Add your project context

Expand your request with the information the assistant should use. For example, you can ask it to review the project's critical path and assess whether the deadline is achievable.

The critical path is the sequence of dependent tasks that determines the earliest possible project finish date.

For this review, ask the assistant to use the project's tasks, recorded dependencies, durations, and scheduled dates, and to flag any missing information:

Use this project's tasks, dependencies, durations, and scheduled dates to identify the critical path. Are we on track to meet the November 10, 2026 deadline? Highlight tasks that could delay completion and flag any missing information.
Vaiz AI chat with a critical-path request and the November 10, 2026 deadline entered in the highlighted message field.

Specify the project or board you want to review and explain what you want to find out. Include relevant details, such as the deadline, current priorities, or tasks that need particular attention. You can also paste information from a brief or meeting notes, or add links to related tasks and documents.

Step 3. Describe the result you need

Specify how the answer should be organized. For the critical-path example, request a table with defined columns and a brief conclusion about the deadline:

Show the critical-path tasks in a table with these columns: Task, Assignee, Dependencies, Planned finish date, and Risk.
Start with a brief conclusion: can we meet the November 10, 2026 deadline? If there isn't enough information to assess this, explain what's missing.
The message field shows the requested table format. The assistant's earlier reply is still asking which project to use.

Clarify which project to use before requesting the table. The example also shows why source checks matter: the participation milestone ends on November 13 in the table, while the reply associates it with November 9-12. Resolve that discrepancy against the project records before using the answer to assess the deadline.

Step 4. Specify details and boundaries

State what the assistant may use and what it should leave unchanged. For a review of existing tasks, specify that it should flag gaps and present findings without changing any records:

Review this project's critical path using only existing tasks and recorded dependencies.
Keep task titles, assignees, priorities, and deadlines unchanged. Flag missing durations or dates instead of estimating them.
Separate confirmed blockers from potential risks. Show your findings without creating, updating, or archiving any tasks.
Screenshot of Vaiz project review showing task dependencies and risks.

The illustrated response identifies overdue work, tasks assigned to a deactivated user account, and missing estimates. It does not establish a verified critical path or confirm that the deadline can be met. Add the missing information before requesting another assessment.

Step 5. Review and refine the response

Check the response against your goal and source material. Verify task references, dates, and requirements, then describe any corrections in a follow-up message.

Here is a separate example: preparing a single brief from two copywriting tasks. Identify the source tasks and ask to review the proposal before any changes:

Combine MR-2963 and MR-2964 into one task titled "Web Summit: write event page copy".
Merge the descriptions into a single brief covering the headline, introduction, CTA paragraph, and button text. Preserve all requirements and relevant links. Show me the proposed merged task before changing the originals.
The assistant finds the two source tasks before preparing a combined brief.

Review the proposed title, board, priority, and description. Check that the brief includes both the introduction and the call-to-action requirements. In this example, the proposal uses the higher priority of the two source tasks.

Screenshot of Vaiz showing a proposed merged task brief for review.

If a requirement is missing or the wording needs work, describe the correction in the same conversation before continuing.

Step 6. Create a task

In this example, create a task to write an article about Vaiz AI. Ask to see the draft before creating the task:

Create a task titled "Write an article about Vaiz AI" in this project.
Description: Write a step-by-step guide covering ready-made prompts, project context, task creation, and project documents. Include practical examples and screenshots.
Leave the assignee and due date unset. Show me the task draft before creating it.

Check the title, project, board, group, milestone, and priority in the draft card. Confirm that the proposed values match your request.

The task is still a draft. Select Create task after reviewing its details.

After creating the task, open it and check its location and details. For more on task fields and activity history, see how to create and manage tasks in Vaiz.

Step 7. Prepare project documents

You can also use Vaiz AI to draft project documents. Specify the document, source material, and sections you need. This example shows a product requirements document (PRD) for a Comment History feature.

To request a draft with this structure:

Using the requirements provided, draft a PRD for a Comment History feature. Include an overview, problem statement, goals, non-goals, user stories, and functional requirements. Flag any unanswered questions.
A Markdown excerpt from the Comment History PRD draft in the AI chat, showing goals, user stories, and functional requirements.

Read the full draft in the chat and check that it accurately reflects the agreed requirements. Ask for corrections or additions, then save the agreed version and link it to the relevant tasks.

The same approach can be used for a project plan or a decision log: provide the brief, tasks, or meeting notes, and specify the sections to include. As decisions change, ask the assistant to update the relevant sections.

See the Vaiz documents guide for document organization, editing, and links to related work.

Step 8. Connect an external assistant if needed

You can also work with Vaiz through an external AI assistant using MCP, the Model Context Protocol. This connects project information to a tool you already use for planning or coding.

Follow the Vaiz MCP setup guide for your client, connect your account, and authorize access. The guide covers Cursor, Claude, Claude Code, and Codex.

Start with a small request:

Read this Vaiz task and summarize its requirements.

Then try a more detailed workflow:

Read the linked specification and prepare an implementation plan. List any requirements that need clarification.

Available actions depend on the connection and permissions. Results saved in Vaiz remain available to teammates working through its boards, tasks, and documents.