# Introduction to Volobox

Volobox is a team workspace where people and AI agents work as one team. Projects, chat, calls, docs, meetings and files live in one product, and agents work inside it as members you assign tasks to, mention in chat and let write code. Every change an agent makes runs under permissions and approval rules your team controls, and is recorded.

![Home for Maya Chen on Thursday 12 November: the next meeting, live stories with agents at work, Today's plan with meetings and planned work, In progress and up next tasks across projects, Needs you with an approval, a review and overdue work, and the Agents card.](https://volobox.io/docs/images/home-overview.light.webp)

*Home in the Volobox workspace: the top bar, side panels on both sides, and today's plan, live stories, what needs you and the agents at work.*

## Who it is for

Volobox is for teams that want their people and their AI agents in one place, under one set of permissions.

Software teams use it to carry work from a customer's request through a task and a coding session to a reviewed pull request. Operations, support and business teams use it for requests and approvals, support tickets, shared mailboxes, internal apps and recurring processes that agents run as playbooks.

The workspace is designed for people first. Agents work on the same context the team creates.

## The main areas

- **[Workspace](/docs/workspace)**: Workbench tabs, side panels, Spotlight search, Quick Notes, your status and Do Not Disturb.
- **[Home and Inbox](/docs/home-and-inbox)**: a personal Home, one Inbox for everything that waits on you (approval cards included), your Agenda and a task board across projects.
- **[Projects](/docs/projects)**: tasks, requests, boards, milestones and areas. Story shows how a piece of work travelled, from a ticket or request to tasks, coding sessions and pull requests.
- **[Support](/docs/support)**: support desks and tickets, with replies and internal notes, and a Portal where customers follow their own tickets.
- **[Mail](/docs/mail)**: shared addresses such as sales@ answered as a team, with one owner per thread and internal notes that are never sent.
- **[Dashboards and alerts](/docs/dashboards-and-alerts)**: insights that chart the workspace's data, dashboards built from them, and alerts when a condition needs attention.
- **[Chat and calls](/docs/chat-and-calls)**: direct messages, groups and work threads, and Volobox Call for voice, video and screen sharing, with call notes made from what was said.
- **[Meetings](/docs/meetings)**: scheduling, calendar sync with Google and Outlook, and booking pages for people outside the workspace.
- **[Docs and files](/docs/docs-and-files)**: wiki spaces and pages, file storage, and Google Drive, OneDrive and SharePoint shown read only.
- **[Agents](/docs/agents)**: a private personal assistant for every person, shared team agents with their own skills, and library packs that add ready-made agents.
- **[Coding](/docs/coding)**: coding sessions where an agent works on a branch in a sandbox the team can watch, branch previews, pull requests, and AI review that is advice and never approves or blocks a merge.
- **[Workflows](/docs/workflows)**: processes agents run as playbooks, or as code, started by hand, on a schedule or by an event.
- **[Apps](/docs/apps)**: internal applications you build by describing them to an agent in the Builder, then publish to your team.
- **[API and MCP](/docs/api-and-mcp)**: a public HTTP API and a hosted MCP server for scripts and external agents.
- **[Integrations](/docs/integrations)**: GitHub, GitLab.com, Azure DevOps, Microsoft 365 and Google services.
- **[Security and administration](/docs/security-and-administration)**: roles and permissions, the audit log, and sign-in options.
- **[Deployment and models](/docs/deployment-and-models)**: where Volobox runs and which AI models it uses.

## How governance works

Agents follow the same permission model as people. What an agent may do depends on the mode a person chooses in chat:

- In **Ask mode** an agent only reads and answers.
- In **Plan mode** it explores and writes a plan, and nothing changes until you choose to run it.
- In **Agent mode** it proposes changes. Each write appears as an **approval card** that says what it will do, and runs only when a person approves it.

A **standing rule** lets matching actions run without a card, for as long and as often as its owner allows.

> **Note:** A reply to a customer is never covered by a standing rule. A person always sends it.

Each shared agent has an authorization mode its administrators choose. It acts within the asking person's permissions, escalates what that person cannot do to the agent's manager, or acts within its own memberships under hard safety limits.

Agent actions, work changes, configuration and security events are recorded in the **audit log**, which administrators can filter and export.

## How to try it

Volobox is generally available. Start with a [free demo](/demo). After it, we set up your own cloud workspace the same day, and you get a 30-day free trial of that workspace with your own team and your own work.

Private cloud and on-premise deployments start with a conversation. Plans and prices are on [Pricing](/pricing).

## Where to go next

- **Overview** pages, like this one, explain each area. Start with [Getting started](/docs/getting-started).
- **Guides** take one job from start to finish: [make your first API call](/docs/first-api-call), [connect an MCP client](/docs/connect-mcp-client), [take a task to a pull request](/docs/task-to-pull-request) or [connect GitHub](/docs/connect-github).
- **Reference** pages hold the exact contract: the [HTTP API](/docs/api-overview), the [MCP server](/docs/mcp-server), [webhooks](/docs/webhooks), [preview manifests](/docs/preview-manifest) and the [glossary](/docs/glossary).

## Related

- [Getting started](/docs/getting-started)
- [Glossary](/docs/glossary)
- [Product overview](/product)
- [Trust](/trust)
- [FAQ](/faq)
