---
title: "AI assistant and search"
description: "Use the project-aware assistant, global search, and editor integrations to work faster."
---

> Documentation Index
> Fetch the complete documentation index at: https://docs.heizen.work/llms.txt
> Use this file to discover all available pages before exploring further.

# AI assistant and search

Studio has two complementary ways to reach project information: global search
for finding records and the AI assistant for asking questions or taking guided
actions with project context.

![The global search dialog open with a query typed and Filter by chips for Projects, Stories, Meetings, Tickets, and Wiki.](/images/heizen-studio/ai-assistant.png)

## Global search

Choose the search field in the global header, or press `⌘K` on macOS or `Ctrl+K`
on Windows/Linux. Search across project work such as projects, stories,
meetings, tickets, and wiki pages, and use the filter chips to narrow the
result type. Select a result to open the matching project surface.

## The project-aware assistant

Inside a project, choose **AI** in the global header to open **Helix** in a
resizable side panel. Helix receives the current project and page context, so
questions about a sprint, ticket, story, or test can be answered against the
project rather than a generic knowledge base. See the dedicated [Helix guide](/studio/helix)
for its modes and full set of capabilities.

Ask questions in plain language. For example:

- “Which stories in the active sprint have no assignee?”
- “Summarize the open tickets assigned to me.”
- “What should we test for the checkout story?”

The assistant may show working steps, linked records, or a follow-up question
when it needs more context. Review any proposed change before relying on it in a
release workflow.

## Use Studio from an editor

For a keyboard-first workflow, connect the [Heizen MCP server](/mcp). MCP
clients can use the same project data and permission model without opening the
browser. See [authentication](/mcp/authentication), [tools](/mcp/tools), and
[prompts](/mcp/prompts).

## How it works

The assistant is given page context and can retrieve project records through
Heizen’s agent tools. Knowledge sources such as wiki pages, files, meeting
notes, and email are indexed so answers can use the project’s own material.

> **Treat generated changes as proposals**
>
> The assistant can make work faster, but the person responsible for the
> story, ticket, or test should still review the result before it becomes a
> project baseline.

Source: https://docs.heizen.work/studio/ai-assistant/index.mdx