# OpenWork Docs > ## Agent Instructions > For agents: install the OpenWork desktop app with `brew install --cask openwork` (macOS) or from https://openworklabs.com/download. Do not run `npx openwork`; the npm package named `openwork` is a different project. > Connect an agent to OpenWork with the MCP Gateway: `claude mcp add --transport http openwork https://api.openworklabs.com/mcp/agent`. Teams sign up at https://app.openworklabs.com?mode=sign-up. > Step-by-step agent guide: https://openworklabs.com/docs/start-here/use-openwork-from-an-ai-agent.md > Connect Slack (any organization admin): https://openworklabs.com/docs/slack/connect-slack.md. OpenWork in Slack (@openwork bot) is private alpha for approved organizations. Setup guide: https://openworklabs.com/docs/slack/overview.md. Agent skill: https://openworklabs.com/.well-known/agent-skills/set-up-openwork-slack/SKILL.md. ## Start here - [Open-source Claude Cowork alternative](https://openworklabs.com/docs/start-here/get-started.md): OpenWork is the open-source Claude Cowork alternative: desktop app, web app, AI Gateway, and MCP Gateway to own your AI stack. Free, no sign-in. - [Downloads](https://openworklabs.com/docs/start-here/downloads.md): Download OpenWork for macOS, Windows, and Linux — stable releases and the rolling alpha channel. - [Use OpenWork from an AI agent](https://openworklabs.com/docs/start-here/use-openwork-from-an-ai-agent.md): Exact commands for Claude Code, Codex, Cursor, ChatGPT, and other agents to install OpenWork, set it up for a team, or connect the OpenWork MCP Gateway. - [Pricing and licensing](https://openworklabs.com/docs/start-here/pricing-and-licensing.md): OpenWork's plans — Free, Team, and Enterprise — and how the MIT + OpenWork EE License split works. - [Stewardship](https://openworklabs.com/docs/start-here/stewardship.md): The promises behind OpenWork's open-core split: what stays open, what can move between tiers, and how the MIT conversion works. ### Move from Claude - [Migrate from Claude Cowork](https://openworklabs.com/docs/start-here/migrate-from-claude-cowork.md): Move your files, skills, plugins, MCP servers, and team setup to OpenWork. - [Import your Claude conversations](https://openworklabs.com/docs/start-here/import-claude-history.md): Bring Claude chats and Claude Code sessions into OpenWork as regular conversations you can read, search and continue. ### Help improve OpenWork - [Missing docs?](https://openworklabs.com/docs/start-here/missing-docs.md): Can't find a page you need? Ask us to write it, or contribute one yourself. - [Languages](https://openworklabs.com/docs/start-here/missing-languages.md): Change the OpenWork app language, see which languages ship today, and help improve or add translations. ## Desktop App ### Get started - [Run your first task](https://openworklabs.com/docs/desktop-app/first-task.md): Install the OpenWork desktop app, pick a folder and a model, and run a first task. No sign-in needed. - [Why sign in?](https://openworklabs.com/docs/desktop-app/why-sign-in.md): What the OpenWork desktop app does without an account, and what signing in to OpenWork Cloud adds. ### Connect services - [Connect your services](https://openworklabs.com/docs/start-here/connect-your-stack/connect-services.md): Sign in to OpenWork and connect the accounts your agent can use - [Add a custom MCP server](https://openworklabs.com/docs/start-here/connect-your-stack/add-an-mcp-server.md): Connect a custom or local MCP server that is not available through OpenWork Connect - [Enable Exa search](https://openworklabs.com/docs/start-here/connect-your-stack/enable-exa-search.md): Give your AI Exa web search in OpenWork, either as a connector for your whole organization or as an MCP server in your own workspace. ### Skills - [Create a skill from chat](https://openworklabs.com/docs/start-here/do-work-with-it/create-a-skill-from-chat.md): Turn a useful OpenWork chat into a reusable skill - [Connectors, skills, and plugins](https://openworklabs.com/docs/start-here/do-work-with-it/skills-plugins-and-mcp.md): Use connectors for systems and plugins for the skills that make a system usable ### Models and providers - [Sign in with ChatGPT](https://openworklabs.com/docs/start-here/connect-your-stack/sign-in-with-chatgpt.md): Use your ChatGPT Plus or Pro subscription as a model provider in the OpenWork desktop app. - [OpenWork vs OpenCode models](https://openworklabs.com/docs/start-here/connect-your-stack/openwork-vs-opencode-models.md): Understand the difference between OpenWork inference and OpenCode Go or Zen models - [Using OpenCode Zen & Go Models](https://openworklabs.com/docs/start-here/connect-your-stack/using-opencode-zen-go-models.md): Connect your OpenCode Go provider to OpenWork Desktop - [Add an Anthropic API key](https://openworklabs.com/docs/start-here/connect-your-stack/add-anthropic-api-key.md): Create an Anthropic API key and paste it into OpenWork to use Claude models. - [Sign in with your own Anthropic subscription](https://openworklabs.com/docs/start-here/connect-your-stack/sign-in-with-your-anthropic-subscription.md): Use your paid Claude plan in the OpenWork desktop app through the Claude Code signed in on your own computer. Local only: it does not work with OpenWork Cloud. - [Add a custom LLM or AI gateway](https://openworklabs.com/docs/start-here/connect-your-stack/add-a-custom-llm.md): Bring your own API key or connect any OpenAI-compatible endpoint (OpenRouter, LiteLLM, Azure OpenAI, your own gateway) to OpenWork. - [Use local models (Ollama, LM Studio)](https://openworklabs.com/docs/start-here/connect-your-stack/use-local-models.md): Run OpenWork with a local model such as Qwen, Llama, or Gemma through Ollama or LM Studio, with no API key and no cloud account. - [Use Gemini and Google Vertex AI](https://openworklabs.com/docs/start-here/connect-your-stack/use-gemini-and-vertex-ai.md): Connect Google Gemini to OpenWork with a Gemini API key, or give your team Gemini and Claude through Google Vertex AI (Agent Platform). ### Do work with it - [Control the browser](https://openworklabs.com/docs/start-here/do-work-with-it/control-the-browser.md): How to use OpenWork to automate the built-in browser - [Use desktop apps](https://openworklabs.com/docs/start-here/do-work-with-it/use-desktop-apps.md): Let your agent work in Mac apps. - [Cross-chat memory](https://openworklabs.com/docs/start-here/do-work-with-it/cross-chat-memory.md): How OpenWork can answer questions about past sessions. - [Work with Excel workbooks](https://openworklabs.com/docs/start-here/do-work-with-it/work-with-excel-workbooks.md): Attach, read, and produce .xlsx workbooks with the agent without leaving OpenWork. - [Organize your chats](https://openworklabs.com/docs/start-here/do-work-with-it/workflows.md): Pin, rename, group, and archive chats in the OpenWork sidebar so you can find ongoing work quickly. - [Automations](https://openworklabs.com/docs/start-here/do-work-with-it/automations.md): Have OpenWork do a task for you on a schedule, on your computer or in the cloud, and see how each run went. ### Use it in the browser - [Open OpenWork Web in your browser](https://openworklabs.com/docs/cloud/run-in-the-cloud/open-cloud-in-browser.md): Open a full OpenWork instance in your browser with nothing to install - [OpenWork Web](https://openworklabs.com/docs/cloud/run-in-the-cloud/shared-workspace.md): Run a full OpenWork workspace in the browser, hosted by OpenWork ### My Library - [My Library](https://openworklabs.com/docs/my-library/overview.md): Everything your team shares with you, and what you make yourself, in one place: connectors, skills, plugins and models. - [Models and your own keys](https://openworklabs.com/docs/my-library/models-and-keys.md): See the AI models your organization gives you, sign in where a provider asks, and add your own provider key that only you can use. - [Share from My Library](https://openworklabs.com/docs/my-library/share.md): Give a connector, skill or plugin you made to teammates or teams, with nothing for them to set up. ### Workbot - [Workbot](https://openworklabs.com/docs/workbot/overview.md): Give your team an assistant in the browser that works in the apps they already use, and choose the model your cloud agents use. - [My bot](https://openworklabs.com/docs/workbot/my-bot.md): A background agent that keeps an eye on your email, calendar and team chat, and messages you in Slack only when something needs you. ### Maintenance - [Back up conversation history](https://openworklabs.com/docs/desktop-app/maintenance/back-up-conversation-history.md): Save your local OpenWork conversation history and the app state needed to find it on Windows, macOS, and Linux. ### Troubleshooting - [Diagnostic prompts](https://openworklabs.com/docs/start-here/troubleshooting/diagnostic-prompts.md): Paste-able prompts that let your agent debug OpenWork issues on its own. ## Integrations - [Integrations](https://openworklabs.com/docs/integrations/overview.md): Use OpenWork Connect in your existing AI client through MCP or a specialized client plugin. - [MCP integration](https://openworklabs.com/docs/integrations/mcp.md): One OpenWork MCP URL brings organization skills and connected tools into OpenCode, Cursor, Codex, Claude Code, and other compatible clients. - [Specialized client plugins](https://openworklabs.com/docs/integrations/client-plugins.md): The OpenCode plugin brings OpenWork AI Gateway models and MCP resources into one client. Cursor and Codex integrations are in progress. ### OpenCode — ready - [Use OpenWork in OpenCode](https://openworklabs.com/docs/model-context-protocol/opencode-plugin.md): Install the opencode-openwork plugin in OpenCode 2 and sign in once to get your organization's AI Gateway models and OpenWork MCP connections on every start, without the desktop app. ### Native integrations - [Native integrations](https://openworklabs.com/docs/integrations/native-integrations.md): Services OpenWork connects to directly, without an MCP server: Google Workspace and Microsoft 365. ## Slack ### Connect Slack - [Connect Slack to OpenWork](https://openworklabs.com/docs/slack/connect-slack.md): Let OpenWork search, read and post in Slack as each person who signs in. An admin sets up your organization's Slack app once in about five minutes; then everyone signs in with their own Slack account. ### OpenWork in Slack (private alpha) - [OpenWork in Slack](https://openworklabs.com/docs/slack/overview.md): Mention @openwork in a Slack channel or DM it to work with your own OpenWork connections and permissions. Private alpha for approved organizations. - [Set up OpenWork in Slack](https://openworklabs.com/docs/slack/set-up-the-slack-app.md): Administrator setup for @openwork in Slack: merge the manifest into your existing Slack app, save the signing secret, install the bot, and choose where replies go. - [Connect your Slack account to @openwork](https://openworklabs.com/docs/slack/connect-your-account.md): Link your own Slack account to OpenWork once, then DM or mention @openwork and confirm it answers. - [Troubleshoot OpenWork in Slack](https://openworklabs.com/docs/slack/troubleshooting.md): Fix @openwork missing from Slack autocomplete, an unverified events URL, connect links that repeat, or a bot that starts but never answers. ## AI Gateway ### Overview - [AI Gateway](https://openworklabs.com/docs/ai-gateway/overview.md): Connect OpenAI, Anthropic, Vertex, Bedrock, Azure, LiteLLM, and compatible catalog providers. Control who uses which model and track cost. - [OpenWork Models](https://openworklabs.com/docs/ai-gateway/openwork-models.md): Hand-picked models for everyone in your organization, with no API keys to manage. Subscribe, turn them on, and track the shared usage that comes with the plan. - [Users and teams](https://openworklabs.com/docs/ai-gateway/users-and-teams.md): See who can use which AI Gateway models, their spend limit and what they spent, and change access for everyone, a team or one person. - [AI Gateway API keys](https://openworklabs.com/docs/ai-gateway/third-party-tools.md): Let members create their own AI Gateway API keys to use your organization's models from 3rd party tools and harnesses, and manage those keys. - [Use Cursor with AI Gateway](https://openworklabs.com/docs/ai-gateway/cursor.md): Point Cursor at your organization's AI Gateway and add your models by name. - [Use Claude Code with AI Gateway](https://openworklabs.com/docs/ai-gateway/claude-code.md): Point Claude Code at your organization's AI Gateway with a project settings file, and use your models, Claude or not. - [Use Claude Desktop with AI Gateway](https://openworklabs.com/docs/ai-gateway/claude-desktop.md): Run the Claude app in Gateway mode on your organization's AI Gateway, with an AI Gateway API key and no Anthropic account. - [Use the Codex CLI with AI Gateway](https://openworklabs.com/docs/ai-gateway/codex.md): Point OpenAI's Codex CLI at your organization's AI Gateway with a provider in ~/.codex/config.toml, and use your models, OpenAI or not. - [Use the Codex app with AI Gateway](https://openworklabs.com/docs/ai-gateway/codex-desktop.md): Run Codex in the ChatGPT desktop app on your organization's models: save your AI Gateway API key in the macOS Keychain and add a provider to ~/.codex/config.toml. - [Provider support](https://openworklabs.com/docs/ai-gateway/provider-support.md): Compare AI Gateway providers, supported credentials and APIs, and the differences between Gateway and legacy providers. - [AI Gateway vs LiteLLM vs Bifrost](https://openworklabs.com/docs/ai-gateway/compare-litellm-bifrost.md): OpenWork AI Gateway vs LiteLLM vs Bifrost: license, SSO, spend tracking, MCP, and team sharing compared, plus how to use them together. - [Use your LiteLLM proxy](https://openworklabs.com/docs/ai-gateway/litellm.md): Connect your own LiteLLM proxy to OpenWork AI Gateway with one organization key, each person's own LiteLLM key, keys OpenWork creates for each person, or Microsoft sign-in for a proxy that only accepts JWTs. - [Use your TrueFoundry AI Gateway](https://openworklabs.com/docs/ai-gateway/truefoundry.md): Connect your TrueFoundry AI Gateway to OpenWork AI Gateway with one TrueFoundry admin key. OpenWork creates a key for each person, so TrueFoundry's teams, budgets and logs apply to them. - [TrueFoundry: Too many tokens](https://openworklabs.com/docs/ai-gateway/truefoundry-too-many-tokens.md): Fix a person shown as Too many tokens on a TrueFoundry provider: raise TrueFoundry's limit on personal access tokens per user, or reset the person's tokens from OpenWork. - [How a prompt reaches a model](https://openworklabs.com/docs/ai-gateway/how-requests-flow.md): How the OpenWork desktop app uses one AI Gateway key for every provider, and how the gateway swaps it for the right provider key, including LiteLLM keys. - [Smart Model Router](https://openworklabs.com/docs/ai-gateway/auto-model-routing.md): Let OpenWork pick the right model and thinking level for every prompt, from the AI Gateway models and tasks you choose. - [Google Agent Platform (Vertex AI)](https://openworklabs.com/docs/ai-gateway/google-agent-platform.md): Set up Gemini and Claude on Google's Gemini Enterprise Agent Platform with your own OAuth app and per-member Google sign-in through OpenWork AI Gateway. - [Amazon Bedrock](https://openworklabs.com/docs/ai-gateway/amazon-bedrock.md): Set up Claude and other models on Amazon Bedrock and Bedrock Mantle through OpenWork AI Gateway, with shared AWS keys or per-member sign-in through AWS IAM Identity Center. - [Microsoft Foundry](https://openworklabs.com/docs/ai-gateway/microsoft-foundry.md): Set up Claude on Microsoft Foundry through OpenWork AI Gateway, with a shared Foundry key or per-member Microsoft Entra ID sign-in. ### Providers for your org - [Adding a managed LLM provider](https://openworklabs.com/docs/cloud/share-with-your-team/managed-llm-provider.md): Share one provider key with your org from OpenWork Cloud. It syncs to members' desktop apps automatically. - [Adding a custom LLM provider](https://openworklabs.com/docs/cloud/share-with-your-team/custom-llm-provider.md): Share an OpenAI-compatible endpoint, such as Azure AI Foundry, LiteLLM, or vLLM, with your org from OpenWork Cloud. - [Per-member LLM credentials](https://openworklabs.com/docs/cloud/share-with-your-team/per-member-llm-credentials.md): Bind a separate write-only provider credential to each organization member - [TrueFoundry: a key for each member](https://openworklabs.com/docs/cloud/share-with-your-team/truefoundry-per-member-keys.md): Provision individual TrueFoundry keys to OpenWork desktop apps with an organization API key. ### Usage and cost - [Overview tab: spend and usage](https://openworklabs.com/docs/ai-gateway/usage-dashboard.md): See what your organization spends on AI through AI Gateway: daily usage and estimated cost for the last 31 days, by model, team and person. - [My Usage: your own spend, limits and model prices](https://openworklabs.com/docs/ai-gateway/my-usage.md): Every member can see their own AI Gateway usage: how close they are to their spend limits, what they used in the past 30 days, and what each model they can use would cost them. - [Counting usage](https://openworklabs.com/docs/ai-gateway/counting-usage.md): How AI Gateway counts reported tokens, estimates cost, and handles queries without usage data. - [Token Costs](https://openworklabs.com/docs/ai-gateway/token-costs.md): How AI Gateway estimates costs per request, adds them to your usage totals, and handles pricing limitations. - [Limits](https://openworklabs.com/docs/ai-gateway/limits.md): How AI Gateway caps usage so your organization stays within budget. - [What uses tokens](https://openworklabs.com/docs/start-here/do-work-with-it/what-uses-tokens.md): Understand when OpenWork model usage starts and how to control it. ### Benchmarks - [Knowledge-work cost benchmark](https://openworklabs.com/docs/benchmarks/overview.md): What it costs an AI agent to keep a customer knowledge base up to date, and what happens to quality when you make it cheaper with a skill, subagents or a model router. - [How the knowledge-work cost benchmark was run](https://openworklabs.com/docs/benchmarks/how-it-was-run.md): The task, the data, the planted facts, the grading rules, the skill and the router behind the knowledge-work cost benchmark. ## MCP Gateway ### Overview - [MCP Gateway](https://openworklabs.com/docs/cloud/run-in-the-cloud/cloud-mcp.md): One hosted MCP server that brings your org's skills, plugins, and connections into Claude Code, Codex, Cursor, or any MCP client. - [Connect OpenWork MCP](https://openworklabs.com/docs/start-here/connect-openwork-mcp.md): Add OpenWork MCP to Claude Code, Codex, Cursor, VS Code, or ChatGPT with one command: https://api.openworklabs.com/mcp/agent. ### Connect a client - [Connect OpenCode](https://openworklabs.com/docs/model-context-protocol/opencode.md): Add OpenWork MCP to OpenCode: add a remote MCP entry for https://api.openworklabs.com/mcp/agent to your OpenCode config, then run opencode mcp auth openwork. - [Connect Claude Code](https://openworklabs.com/docs/model-context-protocol/claude-code.md): Add OpenWork MCP to Claude Code with one command: claude mcp add --transport http openwork https://api.openworklabs.com/mcp/agent, then run /mcp to sign in. - [Connect Codex](https://openworklabs.com/docs/model-context-protocol/codex.md): Add OpenWork MCP to Codex with one command: codex mcp add openwork --url https://api.openworklabs.com/mcp/agent, then codex mcp login openwork. - [Connect Claude Desktop](https://openworklabs.com/docs/model-context-protocol/claude-desktop.md): Add OpenWork MCP to Claude Desktop as a custom connector with the URL https://api.openworklabs.com/mcp/agent, or through mcp-remote. - [Connect ChatGPT](https://openworklabs.com/docs/model-context-protocol/chatgpt.md): Add OpenWork MCP to ChatGPT: paste https://api.openworklabs.com/mcp/agent in Settings > MCP servers and sign in with OAuth. - [Connect Cursor](https://openworklabs.com/docs/model-context-protocol/cursor.md): Add OpenWork MCP to Cursor in one click or by pasting https://api.openworklabs.com/mcp/agent as a remote MCP server, then sign in with OAuth. - [Connect VS Code](https://openworklabs.com/docs/model-context-protocol/vs-code.md): Add OpenWork MCP to VS Code with one command: code --add-mcp with the URL https://api.openworklabs.com/mcp/agent, then start the server and sign in. - [Connect Windsurf](https://openworklabs.com/docs/model-context-protocol/windsurf.md): Add OpenWork MCP to Windsurf through mcp-remote with the URL https://api.openworklabs.com/mcp/agent, then sign in with OAuth. - [Connect Zed](https://openworklabs.com/docs/model-context-protocol/zed.md): Add OpenWork MCP to Zed through mcp-remote with the URL https://api.openworklabs.com/mcp/agent, then sign in with OAuth. - [Connect Gemini CLI](https://openworklabs.com/docs/model-context-protocol/gemini-cli.md): Add OpenWork MCP to Gemini CLI with one command: gemini mcp add --transport http openwork https://api.openworklabs.com/mcp/agent, then /mcp auth openwork. - [Connect OpenWork MCP from a remote machine](https://openworklabs.com/docs/cloud/run-in-the-cloud/connect-openwork-mcp-remote-machine.md): Authenticate OpenWork Connect for OpenCode on a remote host while using a browser on your local machine ### Connections - [Connectors](https://openworklabs.com/docs/cloud/share-with-your-team/connectors.md): Let your team's AI read and act in the apps you already use. Add an app once in Manage › Connectors, choose how people sign in and who can use it. - [Connect your own account](https://openworklabs.com/docs/cloud/share-with-your-team/connect-your-own-account.md): When a connector uses each person's own account, sign in once (or add your key) so your AI works in that app with your permissions. - [Sharing MCP connections with your team](https://openworklabs.com/docs/cloud/share-with-your-team/shared-mcp-connections.md): Add Notion, Linear, Slack, Exa, or any MCP server once in OpenWork Cloud, choose how people sign in, and choose who can use it. - [Connect Microsoft 365](https://openworklabs.com/docs/cloud/share-with-your-team/microsoft-365.md): Set up Outlook, calendars, OneDrive, SharePoint, and Teams for your OpenWork organization - [Connect Google Cloud CLI](https://openworklabs.com/docs/cloud/share-with-your-team/google-cloud-cli.md): Let your team use Google Cloud from OpenWork with their own Google accounts. - [Make your MCP server work with OpenWork](https://openworklabs.com/docs/cloud/share-with-your-team/mcp-server-compatibility.md): What an MCP server's OAuth setup must allow so OpenWork can connect your members: redirect URL, client metadata, and registration. - [Tool Tester](https://openworklabs.com/docs/cloud/share-with-your-team/tool-tester.md): Run any tool a connector offers from Settings › Tool Tester, see exactly what was sent and what came back, and turn tools off for your whole organization. ### Share skills and plugins - [Manage plugins](https://openworklabs.com/docs/cloud/share-with-your-team/plugins.md): Create plugins in Manage › Plugins, import them from GitHub, choose who can use them, and group them into Collections. - [Agents, commands, and hooks](https://openworklabs.com/docs/cloud/share-with-your-team/plugin-agents-commands-hooks.md): What agents, commands and hooks are in an OpenWork plugin, where they come from, and where to see them all. - [GitHub Sources](https://openworklabs.com/docs/cloud/share-with-your-team/github-sources.md): Connect GitHub, add the repositories that hold your team's plugins, and keep OpenWork plugins in step with them. - [OpenWork Connect as an Agent Plugin](https://openworklabs.com/docs/model-context-protocol/agent-plugins.md): Install OpenWork Connect as one portable Agent Plugin, or import compatible Agent Plugins into your organization's Library. - [Collections](https://openworklabs.com/docs/cloud/share-with-your-team/collections.md): Group several plugins for the same audience in OpenWork Cloud, so one access change covers the whole set. - [Publish and copy a skill](https://openworklabs.com/docs/start-here/do-work-with-it/publish-and-copy-a-skill.md): Create a skill in OpenWork Cloud, choose who can use it, and make your own copy of a shared skill without changing the original. - [Share skills and MCP servers with your team](https://openworklabs.com/docs/start-here/do-work-with-it/share-your-setup.md): Give teammates the same skills, plugins, connectors, and models through OpenWork Cloud, without copying files or editing MCP config. ## Artifacts - [Artifacts](https://openworklabs.com/docs/dashboards/overview.md): Create, find, and share interactive artifacts in OpenWork. - [Build an MCP App](https://openworklabs.com/docs/dashboards/build-an-mcp-app.md): Build an MCP App widget that OpenWork renders in chat and on dashboards, and the host protocol OpenWork supports. ## SSO ### Set up your organization - [Get started with OpenWork Cloud](https://openworklabs.com/docs/cloud/get-started.md): Connect the desktop app to your organization for shared skills and plugins, managed connections, models, policies, and hosted workspaces - [Team quickstart: from new org to a team skill](https://openworklabs.com/docs/cloud/team-quickstart.md): Create an OpenWork Cloud organization, invite teammates, publish a skill plugin through a Collection, and use it from the desktop app - [Managing members and RBAC](https://openworklabs.com/docs/cloud/members-and-rbac.md): Manage owners, super admins, admins and members, group them into teams, and decide what they can do with built-in and custom roles ### SSO and SCIM - [Auth & Security](https://openworklabs.com/docs/cloud/auth-and-security.md): Let your team sign in with your identity provider and add or remove people automatically - [Set up OIDC single sign-on](https://openworklabs.com/docs/cloud/sso-oidc.md): Connect your identity provider, verify your email domain, and test sign-in before enabling SSO. - [Google Workspace SAML SSO](https://openworklabs.com/docs/cloud/sso-google-workspace.md): Connect Google Workspace to OpenWork Cloud with SAML - [Google Workspace SCIM provisioning](https://openworklabs.com/docs/cloud/scim-google-workspace.md): Current Google Workspace SCIM support for OpenWork - [Okta SAML SSO](https://openworklabs.com/docs/cloud/sso-okta.md): Connect Okta to OpenWork Cloud or a self-hosted OpenWork deployment with SAML - [Okta SCIM provisioning](https://openworklabs.com/docs/cloud/scim-okta.md): Provision OpenWork members and teams from Okta - [Microsoft Entra SAML SSO](https://openworklabs.com/docs/cloud/sso-microsoft-entra.md): Connect Microsoft Entra ID to OpenWork Cloud with SAML - [Microsoft Entra SCIM provisioning](https://openworklabs.com/docs/cloud/scim-microsoft-entra.md): Provision OpenWork members and teams from Microsoft Entra ID ### Policies and operations - [Desktop policies](https://openworklabs.com/docs/cloud/share-with-your-team/desktop-policies.md): Choose what the OpenWork desktop app lets members and teams do, from OpenWork Cloud. - [Agent permissions](https://openworklabs.com/docs/cloud/share-with-your-team/agent-permissions.md): Choose what agents can do on members' computers: run commands, edit files, open websites, and use local skills and MCP servers. - [Team prompt cards](https://openworklabs.com/docs/cloud/share-with-your-team/team-prompt-cards.md): Show your own suggested prompts to everyone, or different prompts to different teams, in the OpenWork desktop app. - [Prompts for agents](https://openworklabs.com/docs/cloud/share-with-your-team/prompts.md): Give every agent in your members' desktop apps standing instructions, for everyone or for one team. - [Security and operations](https://openworklabs.com/docs/cloud/security-and-operations.md): Review OpenWork Cloud security posture, access governance, and availability expectations - [Enterprise](https://openworklabs.com/docs/cloud/enterprise.md): Deploy OpenWork on your own infrastructure ### Observability - [Analytics](https://openworklabs.com/docs/cloud/analytics.md): See who on your team uses OpenWork, which skills, plugins and connectors they use, and how your workflow runs went, from Observability › Analytics. - [Audit logs](https://openworklabs.com/docs/cloud/audit-logs.md): See who changed what in your OpenWork organization, and when, from Observability › Audit logs. Filter by date, event type, actor, result and origin. ### Organization settings - [Billing](https://openworklabs.com/docs/cloud/billing.md): See your OpenWork plan and how many people it covers, start the Team plan by card, and manage payment details in Settings › Billing. - [API keys](https://openworklabs.com/docs/cloud/api-keys.md): Create an organization API key in Settings › API Keys so scripts and CI can call the OpenWork API as you, without anyone signing in. ## Self-host ### Overview - [Self-host](https://openworklabs.com/docs/start-here/self-host.md): Run OpenWork on your own server or infrastructure. - [Evaluate with Docker Compose](https://openworklabs.com/docs/self-host/evaluate-with-docker-compose.md): Run the OpenWork control plane from published images without cloning or building the repository. ### Deploy to your cloud - [Deploy to your cloud](https://openworklabs.com/docs/self-host/deploy-to-your-cloud/overview.md): Run OpenWork in your AWS, Azure, or Google Cloud account with managed MySQL and either Kubernetes or Cloud Run. - [Deploy on AWS](https://openworklabs.com/docs/self-host/deploy-to-your-cloud/aws.md): Run OpenWork on Amazon EKS or ECS Fargate with Amazon RDS for MySQL. - [Deploy on Azure](https://openworklabs.com/docs/self-host/deploy-to-your-cloud/azure.md): Run OpenWork on Azure Kubernetes Service or Azure Container Apps with Azure Database for MySQL. - [Deploy on Google Cloud](https://openworklabs.com/docs/self-host/deploy-to-your-cloud/google-cloud.md): Run OpenWork on GKE Autopilot or Google Cloud Run with Cloud SQL for MySQL. - [Install from OpenWork Cloud (Deployments)](https://openworklabs.com/docs/self-host/deploy-to-your-cloud/managed-installer.md): Use Manage › Deployments in OpenWork Cloud to install your own OpenWork in your AWS or Google Cloud account, approve every change there, and watch its health. - [Create the first administrator](https://openworklabs.com/docs/self-host/deploy-to-your-cloud/first-administrator.md): Securely claim a new private OpenWork deployment without enabling public signup. ### Plan your network - [Private network deployment](https://openworklabs.com/docs/start-here/private-network-deployment.md): Run OpenWork when laptops have internet and VPN access while Den stays inside a private network. - [Air-gapped deployment](https://openworklabs.com/docs/start-here/air-gapped-deployment.md): Plan OpenWork deployments that limit or remove public internet access. - [Outbound network access](https://openworklabs.com/docs/start-here/outbound-network-access.md): Customer IT allowlist guidance for OpenWork desktop, Den, deployment artifacts, and opt-in providers. - [Certificate trust and proxies](https://openworklabs.com/docs/start-here/certificate-trust-and-proxies.md): Configure private CAs and outbound proxies for desktop, local runtimes, Den, and strict MySQL TLS. ### Install and operate - [Enterprise desktop deployment](https://openworklabs.com/docs/start-here/enterprise-desktop-deployment.md): Deploy the OpenWork Enterprise binary with your existing software distribution, optionally pushing one configuration file through MDM. - [Enable OpenWork Gateway](https://openworklabs.com/docs/self-host/gateway.md): Configure Gateway with matching images, explicit destinations, and shared secrets for admins in every organization. - [Upgrade Gateway safely](https://openworklabs.com/docs/self-host/gateway-upgrade.md): Plan the database and writer cutover before enabling Gateway, and disable management without an unsafe schema downgrade. - [Provision organization configuration](https://openworklabs.com/docs/self-host/declarative-configuration.md): Apply teams, providers, MCP connections, desktop policies, and marketplace metadata from a repeatable script. - [Export and import organization configuration](https://openworklabs.com/docs/self-host/export-import-configuration.md): Download an organization's configuration as one JSON file, review it in git, and import it into the same or another OpenWork through reviewed change requests. - [Installer delivery](https://openworklabs.com/docs/start-here/installer-delivery.md): Choose how the install page delivers standard OpenWork desktop installers. - [Brand an on-prem deployment](https://openworklabs.com/docs/start-here/on-prem-branding.md): Configure an organization name, wordmark, accent color, and desktop icon without depending on a public CDN. - [GitHub connector for Helm](https://openworklabs.com/docs/start-here/github-connector-helm.md): Create a GitHub App and configure the OpenWork EE Helm chart for repository connectors. ### Troubleshoot - [Diagnostics](https://openworklabs.com/docs/self-host/diagnostics.md): Check that desktop apps connect, AI providers are set up correctly, and your installation's services work, from Settings › Diagnostics or one API call. - [Network diagnostics](https://openworklabs.com/docs/start-here/network-diagnostics.md): Diagnose DNS, TLS, proxy, and catalog issues in locked-down OpenWork deployments. - [API (723 pages)](https://openworklabs.com/docs/_llms/api.txt): System, Admin, SCIM, OAuth, Authentication, Bootstrap, Cloud, Deprecated, Users, Automations, Organizations, Invitations, Diagnostics, Managed deployments, API Keys, Workflow Runs, Workflows, Artifacts, Codemode Runs, Desktop Policies, Config blobs, Agent Permissions, Inference, Plugins, Devices, SSO, Sandbox organizations, LLM Providers, Inference Providers, Gateway Usage Limits, Members, Permissions, Capability Sources, Direct uploads, Documents, GitHub, Config Objects, Marketplaces, Connectors, Resources, Teams, Webhooks, Workbot, Workers, Telemetry. ## Roadmap - [Roadmap](https://openworklabs.com/docs/roadmap.md): The roadmap for the OpenWork desktop app, portable capabilities, hosted workspaces, and every surface where work happens. ## Changelog - [Changelog](https://openworklabs.com/docs/changelog.md): OpenWork release notes: new features, improvements, and bug fixes. ## OpenAPI Specs - [openapi](https://api.openworklabs.com/openapi.json?public=1) > The links below point to documentation indexes. Follow each `/_llms/` index recursively until you reach documentation pages. ## Indexes - [API (723 pages)](https://openworklabs.com/docs/_llms/api.txt): System, Admin, SCIM, OAuth, Authentication, Bootstrap, Cloud, Deprecated, Users, Automations, Organizations, Invitations, Diagnostics, Managed deployments, API Keys, Workflow Runs, Workflows, Artifacts, Codemode Runs, Desktop Policies, Config blobs, Agent Permissions, Inference, Plugins, Devices, SSO, Sandbox organizations, LLM Providers, Inference Providers, Gateway Usage Limits, Members, Permissions, Capability Sources, Direct uploads, Documents, GitHub, Config Objects, Marketplaces, Connectors, Resources, Teams, Webhooks, Workbot, Workers, Telemetry. - [API / Admin (65 pages)](https://openworklabs.com/docs/_llms/api/admin.txt): Read managed installer release settings and change history, Find the latest published stable OpenWork version, Save a public managed installer source, Configure managed installer approval policy, Verify an immutable published installer candidate, Pin a verified installer after operator qualification attestation, Explicitly restore environment installer defaults, Get free Auto usage across all organizations, Get observed behavior after the free Auto limit, Get security signals across all organizations, Get refused and failed platform requests for the last 24 hours, Get an organization member's feature overrides, Set an organization member's feature overrides, Slack bot message outcomes, Whether each feature is used and working, across organizations, One feature's use and outcomes by organization, Who has and uses each feature in an organization, Which MCP providers let OpenWork sign people in, Check the preset MCP providers now, Get an organization and a bounded member page, Get organization usage and a bounded member page, Get organization activity, people and model insights, Get bounded platform AI usage, Count platform active users, Get the platform Control Center, Refresh the platform Control Center, List the accounts worth writing to, with who and why, Record a sales follow-up step, Group signups by company email domain, Find organizations with observed app adoption, Read cached deployment service readiness, Queue a deployment service readiness check, Read the cloud runner fleet, Read the cloud runner fleet's history, Watch the cloud runner fleet, Read the cloud computer fleet, Read one cloud computer, Read the cloud computer fleet's history, List sandbox organizations, Create a sandbox organization, List models a new sandbox can get, Replace a sandbox join link, Choose what people who join become, Extend or reopen a sandbox, Set a sandbox's model credit, Add sample Microsoft 365 to a sandbox, Add sample Slack to a sandbox, Close a sandbox, Add a platform admin, Remove a platform admin, Get a user's inference usage, Reset a user's inference usage, Delete a user, Set an organization's plan, Set an organization's free seats, Record an organization DPA decision, Grant or revoke complimentary OpenWork Web access, Get an organization's capability overrides, Set an organization's capability overrides, List features and their state, Change a feature's state, Get a bounded admin user page, Get a bounded admin organization page, Load deferred admin analytics, Get admin overview. - [API / Authentication (30 pages)](https://openworklabs.com/docs/_llms/api/authentication.txt): Check initial administrator bootstrap availability, Verify an initial administrator setup code, List an organization's SSO sign-in options, Resolve deterministic login option, Google or Microsoft OAuth callback for a member inference credential, Get an org's OAuth client configuration for a provider, Save an org's OAuth client for a provider, Check which Google APIs a Google Workspace connector can reach, Begin connecting the calling member's account for a provider, Begin connecting the calling member to Google Workspace, Begin connecting the calling member to Microsoft 365, OAuth callback for a provider, Discover external MCP connection requirements, Stream external MCP connection requirements checks, Review a changed External MCP OAuth issuer, Resolve a free-form query to an MCP server, Manually run a tool from an External MCP Connection, Remove an External MCP Connection by stable key, Edit an External MCP Connection, Remove an External MCP Connection, Disconnect (clear credentials for) an External MCP Connection without removing it, Disconnect the calling member's account for a per-member External MCP Connection, Begin the OAuth handshake for an External MCP Connection, Shared OAuth callback for External MCP Connections, OAuth callback for an External MCP Connection, Read Slack assistant setup, Configure Slack assistant installation, Start Slack bot installation, Complete Slack bot installation, Mint MCP access token. - [API / Users (17 pages)](https://openworklabs.com/docs/_llms/api/users.txt): Get my Get set up checklist, Report a Get set up step this app observed, Get my agent, Change my agent, List my agent's checks, List my feature intro state, Record that I saw, dismissed or acted on a feature intro, Get current user, List current user's organizations, List current user's organization SSO sign-ins, List current user's pending invitations, Send current user the OpenWork desktop download link, Update current user profile, Get current user's prompts, Get current user's desktop config, Get current user's install configuration, Download current user's managed OpenWork desktop. - [API / Automations (27 pages)](https://openworklabs.com/docs/_llms/api/automations.txt): Connect this desktop as an Automation runner, See who can copy your Automation, Share your Automation's definition, List Automation definitions shared with you, Add your own paused copy of a shared Automation, Report whether a desktop runner is connected, List where this member's Automations can run, List Automations, Create an active Automation from an app surface, Create an active OpenWork Cloud Automation, Get an Automation, Archive an Automation, Update an Automation, Activate an Automation, Deactivate an Automation, Run an Automation now, List Automation runs, List the caller's Automation runs in a time range, Inspect an Automation run receipt and execution thread, Cancel an active Automation run, Get the status of remote sessions, Read a remote session, Send a follow-up to a remote session, Answer a remote session's question or permission, Stop a remote session, Attach to a session started on one of the member's computers, List the member's sessions across their computers. - [API / Organizations (50 pages)](https://openworklabs.com/docs/_llms/api/organizations.txt): Get active organization, Delete organization, Update organization, Read minimal cached deployment service readiness, Get recent dashboard activity, Get the member inventory revision, Report features the caller used, Set organization audit capture, List supported audit event types, List retained audit operations, List retained operation events, Read audit retention usage, Export one audit snapshot page, Read an immutable organization brand asset, Upload organization brand assets, List the organization's Workbot, Slack and Automation conversations (content-free), Read what Workbot spent on models over the last 30 days, Get organization install page (download desktop app, install OpenWork), Get public install configuration, Open the install page, List approved web origins, Approve a web origin, Remove an approved web origin, Get the settings for members' agents, Change the settings for members' agents, Get how members' agents are doing, Get the tools members' agents may use on a connection, Test Workbot's default model, Get Workbot settings, Change Workbot settings, Get Workbot settings per team, Set a team's Workbot settings, Get the settings for a team's members' agents, Change the settings for a team's members' agents, Get the tools a team's members' agents may use on a connection, Set Workbot settings for everyone, Get Automation settings, Set Automation settings, Get remote session settings, Set remote session settings, Check whether Workbot has an AI provider, Get default models, List models for the default model, Set the organization default model, Set a team's default model, Get prompts, Set the prompt for everyone, Remove the prompt for everyone, Set a team's prompt, Remove a team's prompt. - [API / Config blobs (18 pages)](https://openworklabs.com/docs/_llms/api/config-blobs.txt): Get the config blob JSON Schema, List config areas, Export configuration, List config change requests, Upload a config change request, Get a config change request, Get one apply run of a config change request, Get a config change request's plan, Download a change request's before-snapshot, Download a change request's answers, Save config change decisions, Save config change secret inputs, Approve a config change request, Apply a config change request, Cancel a config change request, Restore the config before a change request, Get config change settings, Update config change settings. - [API / Devices (12 pages)](https://openworklabs.com/docs/_llms/api/devices.txt): List team computers, Get registered OpenWork instance inventory, Get team computer inventory details, Remove a computer from the team list, Disconnect a member's computer, List the computers I can reach, Register a virtual computer, Report this desktop and how its computer use is doing, Get one of my computers and its recent activity, Remove a computer from my list, Rename a computer I registered, Disconnect one of my computers. - [API / SSO (27 pages)](https://openworklabs.com/docs/_llms/api/sso.txt): Get organization SSO connection, Delete organization SSO connection, List organization SSO connections, Get an SSO connection, Delete an SSO connection, Rename an SSO connection, Check an OIDC issuer or discovery URL, Register organization SAML SSO, Add an organization SAML SSO connection, Replace an SSO connection's configuration with SAML, Register organization OIDC SSO, Add an organization OIDC SSO connection, Replace an SSO connection's configuration with OIDC, Create an organization SSO authentication test, Enable the tested organization SSO configuration, Disable organization SSO, Get organization SAML SP metadata, Request an SSO domain verification token, Verify the organization SSO domain, Create an SSO connection authentication test, Enable a tested SSO connection, Disable an SSO connection, Get an SSO connection's SAML SP metadata, Request an SSO connection's domain verification token, Verify an SSO connection's domains, Start an organization SSO authentication test, Cancel an organization SSO authentication test. - [API / LLM Providers (18 pages)](https://openworklabs.com/docs/_llms/api/llm-providers.txt): Read llm-providers by stable key, Apply llm-providers by stable key, Delete llm-providers by stable key, Read llm-providers by id, Delete organization LLM provider, Update organization LLM provider, Test a custom LLM provider endpoint, List LLM provider catalog, Get LLM provider catalog entry, List organization LLM providers, Create organization LLM provider, Get LLM provider connect payload, Set the calling member's LLM provider credential, Delete the calling member's LLM provider credential, List member credential states for an LLM provider, Set one member's LLM provider credential, Block one member's LLM provider credential, Remove LLM provider access grant. - [API / Inference Providers (66 pages)](https://openworklabs.com/docs/_llms/api/inference-providers.txt): Read the caller's inference inventory revision, Watch the caller's inference inventory, Check which providers on the caller's device can be shared with the organization, Fetch models from a custom OpenAI-compatible provider, Test a custom OpenAI-compatible provider before saving it, Create custom OpenAI-compatible gateway provider, Update custom OpenAI-compatible gateway provider, Read organization Gateway usage by UTC day, List inference gateway providers and model groups for model enablement, List available upstream models for inference gateway provider, Add models to inference gateway provider group, List the caller's member sign-in connections, List organization inference gateway providers, Create inference gateway provider, Get inference gateway provider, Delete inference gateway provider, Update inference gateway provider, Get inference gateway provider connect payload, Test an inference gateway provider before adding it, List configured gateway catalog models, List gateway model groups, Create gateway model group, Delete gateway model group, Update gateway model group, List gateway credential sets, Create gateway credential set, Delete gateway credential set, Update gateway credential set, List gateway access grants, Create gateway access grant, Remove inference provider access grant, Update gateway access grant, Remove inference provider access grant, Begin sign-in for a member inference credential, Check browser readiness for member sign-in, Continue member Google or Microsoft sign-in in a signed-in browser, Start member AWS sign-in from a signed-in browser, Check a member AWS sign-in from a signed-in browser, Disconnect the caller's member sign-in for an inference provider, Create LiteLLM gateway provider, Sync LiteLLM models and groups, Update LiteLLM key or key creation, Connect the caller's LiteLLM key, Connect a LiteLLM key from a signed-in browser, Retry creating the caller's LiteLLM keys from a signed-in browser, Create TrueFoundry gateway provider, Sync TrueFoundry models and people, Update TrueFoundry admin key, missing users, usage costs or models, Reset a person's TrueFoundry tokens, Retry creating the caller's TrueFoundry key from a signed-in browser, Move an LLM provider to the inference gateway, Get AI Gateway settings, Set AI Gateway settings, List my AI Gateway API keys, Create an AI Gateway API key, Revoke my AI Gateway API key, List the organization's AI Gateway API keys, Revoke a member's AI Gateway API key, List Smart Routers, Create a Smart Router, List models for Smart Routers, Update a Smart Router, Delete a Smart Router, Set who a Smart Router applies to, List my Smart Routers, Connect to Smart. - [API / Gateway Usage Limits (17 pages)](https://openworklabs.com/docs/_llms/api/gateway-usage-limits.txt): Read own Gateway usage by UTC day, List usage limit policies, Create usage limit policy, Update usage limit policy, Archive usage limit policy, Restore archived usage limit policy, List policy assignments, Assign usage limit policy, Remove policy assignment, Search members for usage limits, Read own Gateway usage limits, Inspect member Gateway usage limits, List organization usage increase requests, Request usage limit extension, List own usage increase requests, Approve 25 percent usage extension, Deny usage extension. - [API / Capability Sources (86 pages)](https://openworklabs.com/docs/_llms/api/capability-sources.txt): Check whether the calling member has connected a provider, Disconnect the calling member's account for a provider, Prepare a resumable Google Drive upload, or a replacement of an existing file's content, for a client with local file transport, Create a five-minute download link for one Google Drive file, for clients with a shell, Send an existing Gmail draft, List or search Gmail drafts, Create a Gmail draft or threaded reply with optional workspace attachments, Read a Gmail draft, Replace a Gmail draft's full content, Permanently delete a Gmail draft, Archive, unarchive, mark read or unread, star or change Gmail message labels, Move a Gmail message to trash, Restore a Gmail message from trash, List Gmail labels, Create a Gmail user label, Permanently delete a Gmail user label, Update a Gmail user label, Read an individual Google Calendar event on a selected calendar, Delete/cancel a Google Calendar event after explicit confirmation, Update or reschedule a Google Calendar event or recurring meeting; use the master ID to change the whole series or recurrence, Read Google Sheets spreadsheet metadata and sheet names without grid data, Create a real Google Sheets spreadsheet with one bounded initial sheet, Read a bounded rectangle of Google Sheets cells, Write Google Sheets cells, RAW by default; USER_ENTERED requires explicit approval, Append rows after a Google Sheets table found in a bounded range; inserts rows, RAW by default, Create a Google Drive folder, optionally in a selected parent, Read Google Drive file metadata, including current parent IDs for a move, Rename, move, trash, or restore Google Drive file metadata; never permanently delete, Inspect Google Docs tabs and headings before reading or editing passages, Read or search a small Google Docs passage without loading the whole document into model context, Replace an exact saved Google Docs passage only at the revision the member read, List Google Chat spaces, group chats and direct messages, Find the Google Chat direct message with a person, Read recent messages in a Google Chat space, Send a message to an existing Google Chat space or direct message, List or search Gmail messages as the calling member, Read a Gmail message with its plain-text body as the calling member, Download a Gmail attachment's bytes as the calling member, List Google Calendar events in a time range as the calling member, Create a Google Calendar event (one-time or recurring series) as the calling member, Add a Google Meet link to a Calendar event, Search Google Drive files as the calling member, Read a Google Drive file's text or binary content as the calling member, Share a Google Drive file with a person or the organization, List a OneDrive or SharePoint folder, Find the OneDrive, SharePoint or Teams file behind a link, Search SharePoint sites, List a SharePoint site's document libraries, Create a five-minute download link for one OneDrive or SharePoint file, for clients with a shell, Prepare an upload to OneDrive or SharePoint, or a replacement of an existing file's content, for a client with a shell, Send an existing Outlook draft after explicit user confirmation, Create a reply draft for a selected Outlook message, List an Outlook message's attachments as the calling member, Download an Outlook attachment's bytes as the calling member, Read an Outlook message as the calling member, Update Outlook message read state or categories, Move an Outlook message to archive, deleted items, or inbox, Delete a selected Outlook calendar event, Edit or reschedule an Outlook calendar event, Cancel an Outlook meeting and notify attendees, Read a OneDrive or SharePoint text or binary file as the calling member, Rename or move a OneDrive or SharePoint item within the same drive, Create a folder in OneDrive or a SharePoint library, List or search Outlook mail as the calling member, List Microsoft 365 calendar events as the calling member, Create an Outlook calendar event as the calling member, Search OneDrive and SharePoint files as the calling member, Create or replace a OneDrive text file as the calling member, Create an Outlook draft as the calling member, List Microsoft Teams chats as the calling member, List messages in a Microsoft Teams chat as the calling member, Send a message to an existing Microsoft Teams chat as the calling member, List predefined External MCP Connection presets, List External MCP Connections, Register a new External MCP Connection for the org, Get the tool policy for an External MCP Connection, Update the tool policy for an External MCP Connection, Check connectors that have no stored status, Check and record connection readiness, List tools exposed by an External MCP Connection, List MCP Apps exposed by an External MCP Connection, Create or replace an External MCP Connection by stable key, Get an External MCP Connection, Replace who can use an External MCP Connection, Prepare removing an organization connection, Ask a person to enter the API key of an existing MCP connection. - [API / Documents (17 pages)](https://openworklabs.com/docs/_llms/api/documents.txt): Resolve an authorized Drive document link without downloading content, Open a Drive document in the inline editor, Save explicit inline edits to the same Drive file, Preview a Drive PDF or image, or play a video or audio file, Open a document, spreadsheet or presentation editor, Check whether an Office save is ready, Save an Office snapshot to Drive, Discard the temporary Office session without changing Drive, Read metadata for one temporary editor session, Manage a temporary editor session lock, Read one temporary editor file, Stage editor changes without writing to Drive, Browse, search and list recent OneDrive files, Read a OneDrive file's details without its content, Open a OneDrive file in the inline editor, Save explicit inline edits to the same OneDrive file, Preview a OneDrive PDF or image, or play a video or audio file. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.