Skip to main content
Macro’s AI Chat gives you a powerful assistant that understands your entire workspace. Ask questions about your emails, documents, messages, and more - or let the AI search autonomously with agent mode.

Key Features

Context-Aware

AI understands your workspace and can reference your content

@Mention for Context

Attach documents, emails, or files to give AI specific context

Agent Mode

AI can autonomously search and explore your workspace

Web Search

Query both your workspace and the broader web

Getting Started with AI Chat

1

Open AI Chat

Create a new chat or open an existing conversation
2

Ask a question

Type your question or request in natural language
3

Add context (optional)

@mention documents, emails, or files for specific context
4

Get AI response

AI responds with information from your workspace or the web
AI Chat works best when you give it context. @mention relevant documents or describe what you’re working on.

What AI Chat Can Do

Answer Questions About Your Content

Ask about information in your workspace:
  • “What did the team decide about the API redesign?”
  • “Summarize the customer feedback from last week”
  • “What are my tasks due this week?”
  • “Show me emails about the Q4 budget”

Summarize Documents

Get quick summaries of long content:
  • “Summarize @quarterly-report”
  • “What are the key points in @meeting-notes?”
  • “Give me a brief overview of @design-doc”

Search Your Workspace

Let AI search for you:
  • “Find documents about the authentication system”
  • “What tasks are assigned to me?”
  • “Show me recent channel messages about the launch”
Get information from the web:
  • “What are best practices for API rate limiting?”
  • “Research competitors in the project management space”
  • “Find recent articles about CRDT technology”

Draft and Compose

Get help writing:
  • “Draft an email to the customer about the delay”
  • “Help me write a project proposal”
  • “Suggest improvements to this paragraph”

Using @Mentions for Context

One of the most powerful features is the ability to @mention content for context:

How to Add Context

  1. Type @ in the chat input
  2. Search for and select documents, emails, files, or channels
  3. Selected items appear as attachments
  4. AI uses these attachments as context for its response

What You Can @Mention

  • Documents: Give AI access to specific docs
  • Emails: Reference email threads
  • Files: Attach PDFs, images, or other files
  • Channels: Reference channel conversations
  • Tasks: Provide task context
  • Open Tabs: Use @open-tabs to reference all your currently open documents
When you @mention a document in AI chat, the AI reads and understands its full content. This is perfect for asking questions about specific documents.

Example: Document Analysis

User: @quarterly-report What were our revenue numbers for Q3?

AI: Based on the quarterly report you attached, Q3 revenue was 
$2.4M, representing 18% growth over Q2...

Example: Multi-Document Context

User: Compare the approaches in @design-doc-v1 and @design-doc-v2

AI: Looking at both documents, here are the key differences:
1. V1 proposes a microservices architecture while V2 suggests...

Agent Mode

Agent mode allows AI to autonomously search and explore your workspace:

How Agent Mode Works

1

Enable agent mode

Toggle agent mode in the chat settings
2

Ask your question

AI will autonomously decide what to search for
3

Watch AI search

See what the AI is searching and finding in real-time
4

Review results

AI synthesizes findings into a comprehensive answer

When to Use Agent Mode

  • Broad questions: “What’s the status of the redesign project?”
  • Multi-source research: “Compile all customer feedback on the new feature”
  • Discovery: “What have we discussed about performance optimization?”
  • Synthesis: “Summarize our approach to user authentication across all docs”
Agent mode is powerful for exploratory questions where you’re not sure exactly which documents or emails contain the information you need.

Chat Features

Message History

All conversations are saved:
  • Return to previous chats anytime
  • Continue conversations across sessions
  • Reference earlier messages in new questions
  • Build on previous context

Edit Messages

Refine your questions:
  • Edit previous messages to rephrase questions
  • AI will regenerate responses based on edits
  • Keep conversations clean and focused

Attachments

Attach context to your questions:
  • Documents from your workspace
  • Email threads
  • Files and PDFs
  • Channel conversations
  • Multiple attachments per message

Model Selection

Choose the right AI model:
  • Different models for different tasks
  • Balance speed vs. capability
  • Model settings persist per chat

Best Practices

Be Specific

The more specific your question, the better the response: ❌ “Tell me about the project” ✅ “What are the key milestones in @project-roadmap and their deadlines?”

Provide Context

Use @mentions to give AI the right context: ❌ “What did we decide?” ✅ “What did we decide in @meeting-notes about the API design?”

Break Down Complex Questions

For complex requests, break them into steps:
  1. “Summarize @design-doc”
  2. “What are the main technical challenges?”
  3. “How do these compare to @alternative-approach?”

Use Follow-ups

Build on AI responses with follow-up questions:
  1. “What are my tasks this week?”
  2. “Show me more details about the API task”
  3. “Who else is working on this?”
AI Chat remembers the conversation context, so you can ask follow-up questions without repeating context.

Use Cases

Project Status Updates

What's the current status of the authentication project? Include 
recent tasks, channel discussions, and any related documents.

Customer Research

Summarize all customer feedback about the mobile app from the last 
month. Include emails, channel messages, and support docs.

Meeting Preparation

@meeting-agenda Help me prepare for tomorrow's meeting. What are 
the key topics and what background should I review?

Document Analysis

@contract Review this contract and highlight any unusual terms or 
potential concerns.

Research and Learning

I'm working on @api-redesign. Research best practices for RESTful 
API design and versioning strategies.

Privacy and Security

What AI Can Access

  • Only content you have permission to view
  • Only content you explicitly @mention or that agent mode finds
  • Your AI chats are private to you

What AI Cannot Access

  • Content you don’t have permission to view
  • Other users’ private chats
  • Content outside your workspace (unless web search is enabled)
AI respects all Macro permissions. It can only access and reference content that you have permission to view.

Integration with Macro

AI uses Macro’s search to find relevant content:
  • Agent mode searches your workspace automatically
  • Results are synthesized into natural responses
  • You can see what AI found and searched

With Documents

Create content from AI responses:
  • Save AI responses as documents
  • Reference AI chats from documents
  • Use AI to draft document sections

With Tasks

Manage tasks with AI:
  • Ask about your tasks and deadlines
  • Get task summaries and status updates
  • Create tasks from AI conversation
AI understands your knowledge graph:
  • Follows @links in documents
  • Discovers related content
  • Builds on your connected information

Keyboard Shortcuts

ActionShortcutDescription
Focus InputEnterFocus the chat input field
Send MessageCmd+EnterSend your message
New ChatCmd+NStart a new conversation
@ Mention@Open mentions menu for context

Tips for Effective AI Chat

  1. Start specific questions with @mentions to relevant documents
  2. Use agent mode for broad, exploratory questions
  3. Ask follow-up questions to dive deeper
  4. Edit and refine your questions if the response isn’t quite right
  5. Attach multiple documents for comparative analysis
  6. Use natural language - AI understands conversational queries

Build docs developers (and LLMs) love