# Executor docs > Connect your accounts once, then use them from the dashboard or any MCP client. ## When to use Executor Executor holds the credentials for services you already use and turns them into tools. Reach for it when an agent needs to call a real service without holding the credential: connect an account once, then call the app through one MCP endpoint at `/mcp`. The hosted origin is `https://v2.executor.sh`. Executor also runs locally and self-hosted. ## Getting started - [What Executor is](https://v2.executor.sh/docs): Executor holds the services you already use as reusable accounts, connects them to apps, and gives both people and agents one place to call them. - [Connect an account](https://v2.executor.sh/docs/connect-an-account): Save a credential once as an account, from the dashboard, the app you are configuring, or an agent, then select it for any app that needs that provider. ## Agents and MCP - [The MCP endpoint](https://v2.executor.sh/docs/mcp): One endpoint gives an agent every app you have configured, without putting a credential in the agent. Codemode, sign-in and elicitation modes explained. - [Add an MCP client](https://v2.executor.sh/docs/mcp-clients): Connect Claude Code, Cursor, Codex or any other MCP client to your Executor endpoint, on hosted, self-host or a local server with an API key. ## Concepts - [Providers and accounts](https://v2.executor.sh/docs/concepts/providers-and-accounts): A provider describes how to authenticate with a service. An account is one saved, reusable instance of it, and an app requirement is the slot it fills. - [Apps and deployments](https://v2.executor.sh/docs/concepts/apps-and-deployments): An app is one configured copy of deployed code, with its own account selections. A deployment is one immutable, retained version of that code. - [Tools and approvals](https://v2.executor.sh/docs/concepts/tools-and-approvals): An app's queries and mutations become tools automatically. Each one declares its own approval, and nothing is inferred from an operation looking read-only. - [Organizations and access](https://v2.executor.sh/docs/concepts/organizations-and-access): Hosted and self-host group people into organizations. An organization owns its accounts, apps and deployments, and each member holds one role. ## Build - [Author an app](https://v2.executor.sh/docs/build/author-an-app): Write an app in plain TypeScript: a provider, some queries and mutations, the approvals that guard them, stored data, and the call that deploys it. ## Run Executor yourself - [Command line](https://v2.executor.sh/docs/run/cli): Run Executor on your own machine for one person. Your accounts, apps and data stay there, behind an API key, with an MCP endpoint any local agent can use. - [Self-host with Docker](https://v2.executor.sh/docs/run/self-host): Run the whole hosted server in one container, with an embedded database and no separate Postgres, queue or proxy, on your own infrastructure. - [Tracing](https://v2.executor.sh/docs/run/tracing): Read a slow request as a waterfall. Executor bundles a local trace store and viewer, and exports OTLP to any collector you point it at instead. ## Agent resources - [llms-full.txt](https://v2.executor.sh/docs/llms-full.txt): The full Markdown of every page in one file. - [Page Markdown](https://v2.executor.sh/docs/index.md): Append `.md` to any page URL to fetch that page as raw Markdown. - [JSON API](https://v2.executor.sh/docs/api/docs/pages.json): Page index of the JSON docs API; each entry links the page's JSON and Markdown forms. Described by the OpenAPI document at https://v2.executor.sh/docs/openapi.json. - [API catalog](https://v2.executor.sh/docs/.well-known/api-catalog): RFC 9727 linkset of the APIs documented here. - [AI catalog](https://v2.executor.sh/docs/.well-known/ai-catalog.json): ARD manifest of the agent-facing resources on this site (MCP server, skills, APIs). - [agent-readability.json](https://v2.executor.sh/docs/agent-readability.json): Manifest of every agent-facing artifact on this site. - [Sitemap](https://v2.executor.sh/docs/sitemap.xml): Every indexable page URL with its last-modified date.