Runwork
Connect · AI Tools

Your AI tools
join the team

Connect any AI tool to your team's cloud and it becomes part of the team. What it learns there stays with the team.

Get Runwork Start free trial MCP docs
Connect People & Teams AI Tools Integrations MCP Community

60+ AI tools, connected with no setup

One command, runwork setup, finds the AI tools on each computer and connects them to your workspace. Anything that supports MCP connects too.

ChatGPT ChatGPT Full support
Claude Desktop Claude Desktop Full support
Claude Code Claude Code Full support
Codex Codex Full support
Cursor Cursor Full support
Windsurf Windsurf Full support
Gemini CLI Gemini CLI Full support
Antigravity Antigravity Full support
Plus any MCP-compatible tool, yours included

Full support means runwork setup wires up the desktop app, CLI, local agent config, skills, and MCP automatically. Any MCP-compatible tool can also connect manually using your workspace endpoint URL and API key.

What's in your work cloud

Everything your agent gets access to the moment it connects.

3,300+
integrations
callable as tools
60+
AI tools
auto-configured
∞
team-built skills
+ skills.sh + MCP Registry

Plus your team's own entities, workflows, agents, components, and apps. The agent sees the same surface area your humans do.

How agents connect

MCP (Model Context Protocol)

Every workspace exposes an MCP server. Your AI agent connects once and gets tools for everything: entities, workflows, schedules, integrations, files, skills.

Streamable HTTP and SSE transports. Standard protocol. Works with any MCP-compatible tool.

Learn more

Runwork CLI

Full terminal interface to the workspace. Every command supports --json for machine-readable output that AI agents parse natively.

runwork agents, runwork entities, runwork workflows, runwork integrations, runwork files, and more.

Learn more

Skills (SKILL.md)

Every app auto-generates a skill file with capability descriptions, API references, and domain knowledge. Your agent reads it and understands what's available.

YAML frontmatter + markdown. Portable to any AI tool. Downloaded from dashboard or synced via CLI.

Learn more

Knowledge files (CLAUDE.md, AGENTS.md)

When any AI agent creates or opens a Runwork app (via runwork init or runwork dev), the platform generates knowledge files on the fly. This works for any agent: Claude Cowork on the desktop creating a workflow, Claude Code building a full app, or Cursor adding an endpoint. The agent understands Runwork from the first prompt.

Created automatically. Always up to date. Works with Claude Cowork, Claude Code, Cursor, Codex, and any agent that reads project files.

Learn more

Community sync (skills.sh + MCP Registry)

Pull skills and MCP servers from public sources directly into your workspace. Your agent can install community-built tools from skills.sh + MCP Registry the same way it loads team-authored ones.

Installable via CLI or dashboard, isolated per workspace.

Learn more

What your AI tools do here

Three modes once connected: your agent knows the business, operates the workspace, and makes things the team keeps.

Know

Work with full context of your business.

  • Full workspace context

    Your agent sees your data, connected integrations, and team skills from the first prompt, with no pasting background into every chat.

  • Tap team skills

    Use skills your teammates have built. What one person teaches their agent, every agent on the team can do.

  • Hand off conversations

    Pass an AI conversation to a teammate; they resume it natively in their own agent, with the full thread and context intact.

  • Persistent memory

    Conversations, prior apps you built together, and team context carry across sessions. Your agent picks up where it left off.

Do

Operate the workspace through the MCP server.

  • Query and write entities

    Read and update your team's data model: customers, projects, leads, vendors, anything you've defined.

  • Call 3,300+ integrations

    HubSpot, Stripe, Slack, Google, Notion, Linear. Every connected integration becomes a tool the agent can call directly.

  • Trigger workflows & schedules

    Run multi-step automations with parameters or kick off scheduled jobs on demand.

  • Files & storage

    Upload reports, download data, manage workspace buckets the same way humans do.

  • Call public endpoints

    Hit any HTTP endpoint your team has exposed from a workspace app: webhooks, APIs, internal tools.

Build

Create new apps, skills, and agents from conversation.

  • Vibe-code full apps

    Spin up a full-stack app from a prompt: runwork init → runwork dev → runwork deploy. Frontend, backend, schema, integrations.

  • Author skills & agents

    Write team-shared skills, register MCP servers, deploy task agents. Anything you build syncs to every teammate's tools.

  • Drop in components

    Reusable UI building blocks for the apps your agent assembles. No styling guesswork.

  • Delegate to agent sandboxes

    Hand off heavy or long-running work to isolated compute environments with full code execution.

Permissions, audit logs, and adoption

Your AI tools follow the same rules as people. Each one gets a role, permissions decide what it can reach, and everything it does lands in the audit log. The adoption dashboard shows the work your AI actually does for the team.

Permissions

Roles decide what each person, and each person's AI, can touch.

Audit logs

Every action an AI tool takes in the workspace is recorded.

Adoption

A dashboard that shows who's improving and what their AI gets done.

Connect in 60 seconds

The easy way

Download the desktop app

It signs into your workspace, detects the AI tools on the computer, connects them to your team cloud, and quietly keeps everything in sync from the menu bar. Nothing to configure, so it works for every teammate, technical or not.

1. Download 2. Sign in to your workspace 3. Your agents are detected and synced
Download the desktop app Mac, Windows, and Linux

Prefer the terminal? Same setup, three commands.

Terminal
# Install the CLI
$ curl -fsSL https://runwork.ai/install.sh | sh
# Authenticate against your team workspace
$ runwork login
# Connect every local AI agent to the team work cloud
$ runwork setup
Detected agents:
✓ Claude Code
✓ Claude Desktop
✓ Cursor
✓ Codex
Syncing from team workspace…
✓ 23 team skills installed
✓ Workspace MCP server connected
✓ 12 agents available
✓ Integrations connected: hubspot · gmail · slack
✓ Your agents joined the team work cloud.

That's it. Every connected agent now sees the same skills, data, integrations, and automations your team uses, and stays in sync as the workspace evolves.

The Desktop app runs the same setup with a guided UI, so every teammate gets there without touching a terminal.

Questions people actually ask

Mostly about connecting what you already have.

Which AI tools work with our existing stack of Stripe, HubSpot and Slack?
All of them, because the stack is connected to the workspace rather than to any one AI tool. You connect Stripe, HubSpot or Slack once, and every teammate's AI can use them from then on: 3,300+ integrations are available and each connected one becomes a tool your agent can call directly. ChatGPT, Claude, Cursor, Codex, Windsurf and Gemini are set up automatically, and 60+ AI tools are in the registry. The person who connects HubSpot does it once, for everyone, not once per teammate and not once per tool.
Can our AI actually do things in those apps, or only read from them?
Both. A connected integration is a tool the agent can call, so it reads records and writes them: create the deal, send the message, update the row, move the file. It runs as your workspace with the permissions your role allows, so it can do what you could do and not more, and what it did is written into the record afterwards.
Does connecting all this need a developer?
No. One command, runwork setup, finds the AI tools already installed on a computer and connects them, and the desktop app runs the same thing with a guided screen for anyone who would rather not open a terminal. Connecting an integration is an authorisation screen, the same as connecting any app to any other.
Do we all have to standardise on the same AI tool?
No, and we would argue against trying. People keep whatever they already chose. What the team builds, the skills, data, integrations and automations, is kept in the workspace instead of inside one company's product, so somebody switching from ChatGPT to Claude next quarter is their decision rather than your migration project.
Can a teammate pick up an AI conversation I started, instead of me pasting the transcript?
Yes. Hand off the conversation and your teammate resumes it in their own AI tool, with the thread and the context intact. It does not have to be the same tool you were using. That is the point of it: the work moves between people without also having to move between products, and nobody reconstructs a chat from a screenshot.
What does an AI tool know about our business the first time it connects?
Your workspace: the data your team keeps, the integrations you have connected, and the skills your teammates have saved. It reads that from the first prompt, so nobody starts a session by pasting in background. Anything the team adds later shows up without each person reconnecting.
Can we see what everyone's AI has been doing?
You can see what it did, which is a narrower thing than what it said, and the distinction is worth being exact about. Actions in the workspace are recorded: creates, updates and deletes on your data, permission and team changes, sign-ins, API calls and automation runs, each with who did it, when, and what the value was before. Conversations held inside somebody's AI tool are not recorded, so this answers what was done to your business and not what an assistant suggested in a chat.
What if the tool we use is not on your list?
Every workspace exposes an MCP server, so anything that speaks MCP connects with a workspace URL and an API key, including something you wrote yourself. The named tools get the extra convenience of being configured automatically; they do not get access that others cannot have.

Connect your AI tools

Every AI tool your team uses, in one cloud.

Get Runwork Start free trial
Ask your own AI about this