> ## Documentation Index > Fetch the complete documentation index at: https://archie.com/docs/llms.txt > Use this file to discover all available pages before exploring further. # MCP Server > Enable a per-project MCP Server so your project's users and agents can query its data from an AI assistant. It is data-only, inherits each user's project roles, and you configure its identity and tools. The **MCP Server** is a per-project connector for your project's own data. Once enabled, your users and their AI agents connect from a tool like Claude, Cursor, or ChatGPT and work with that project's data — governed by each user's project role, and limited to a **data-only** set of tools. It is separate from the [Archie MCP](https://archie.com/docs/features/backend/app-services/mcp-server/archie-mcp): the Archie MCP manages the platform; this one is scoped to a single project. ## Enable it **1. Open the panel** Go to **Backend → App Services → MCP Server**. **2. Turn it on** Toggle the MCP Server **on**. The connection details and configuration appear below the toggle. > [!NOTE] > > Enabling the MCP Server does **not** open your data to everyone. Every request still authenticates as a specific user and runs with that user's [project roles](https://archie.com/docs/features/backend/app-services/role-based-access.md). Enabling it only makes the connector *available*; access is still governed by your permissions. ## Server identity Give the server a **name, icon, and description**. These are optional — leave them blank to use the defaults. | Field | What it does | Notes | | - | - | - | | **Name** | The name assistants show for this connector. | Shown by most clients (for example, as the connector's title). | | **Icon** | An emoji or an `https` URL to a logo image. | Used in Archie's UI and, when you publish to an assistant's directory, as the listing logo. Assistants that add the server by URL usually show a letter avatar regardless. | | **Description** | A short description of what this server gives access to. | Sent to the assistant as context (it shapes how the model uses the server); it is not shown as a label in every client's UI. | Edit the fields and click **Save**. ## Authentication How users connect depends on the [authentication providers](https://archie.com/docs/features/backend/app-services/authentication-providers/overview.md) your project has enabled. The panel adapts the connection instructions to match: **Sign-in (identity)** If your project uses an identity provider (Archie Auth, Google, Microsoft, and so on), connecting **opens a login window**. The user signs in with their project account — there is no token to paste. After sign-in, the assistant acts with that user's project role. **API key** If your project uses API keys, the user authenticates with a **project API key** sent as an `Authorization: Bearer` header. Create keys under **Backend → App Services → Authentication Providers**. The connection config the panel shows already includes the header — the user just replaces the placeholder with their key. > [!WARNING] > > If no authentication provider is active on the project, users can't connect. Enable at least one under [Authentication Providers](https://archie.com/docs/features/backend/app-services/authentication-providers/overview.md) first. ## Tools The MCP Server exposes a **data-only** tool set. Assistants use these to understand your API and work with your data — never to administer the platform. Clients group them as **read-only** or **write**: | Tool | Type | What it does | | - | - | - | | `list_operations` | Read | Lists the queries and mutations in your project's GraphQL API (schema introspection). | | `get_field_details` | Read | Returns the fields and arguments of a type, so the assistant can build a valid query. | | `semantic_search` | Read | Searches your data by meaning, when [vector search](https://archie.com/docs/features/backend/data-model/data-viewer.md) is enabled. | | `read_archie_docs` | Read | Looks up Archie product documentation. | | `get_context` | Read | Returns the current project and environment scope for the session. | | `execute_graphql_operation` | Write | Runs a GraphQL query or mutation against your project — governed by the user's role and `ALLOW_MUTATIONS`. | | `set_context` | Write | Session-local scope helper. | > [!NOTE] > > The exposed set is the intersection of this built-in floor and your own selection: you can **turn a tool off** to withhold it, but you can't add tools beyond the data-only floor. Administrative and platform tools are never available here. To narrow the set, toggle individual tools off in the panel. An empty selection uses the full data-only floor. ## Permissions The MCP Server never widens access. A connected user acts with their **own** project role: * [Role-Based Access](https://archie.com/docs/features/backend/app-services/role-based-access.md) decides which operations and tables the user can touch. * [Row-level filters](https://archie.com/docs/features/backend/app-services/row-level-filters.md) apply exactly as they do on the GraphQL API. * Mutations run only when the user's role allows the write **and** the environment permits mutations. This is the same authorization the GraphQL API enforces — the MCP is just another authenticated caller. ## Next step Ready to connect a tool? See [Connecting an assistant](https://archie.com/docs/features/backend/app-services/mcp-server/connecting). ## FAQ > **Does enabling the MCP Server expose all my data?** > > No. Every request authenticates as a specific user and runs with that user's project role and row-level filters. The connector is data-only and never exposes platform or admin operations. > **Is this the same as the Archie MCP?** > > No. The [Archie MCP](https://archie.com/docs/features/backend/app-services/mcp-server/archie-mcp) manages the platform across all your projects. The MCP Server is scoped to one project and works with that project's data. > **Can users write data, or only read it?** > > Both, within their permissions. Writes go through `execute_graphql_operation` and are allowed only when the user's role permits the mutation and the environment allows mutations. Read-only and write tools are grouped separately so clients can require approval for writes. > **Why doesn't my icon show up in the assistant?** > > Most assistants show a letter avatar for a connector added by URL — the per-server icon mainly matters when you publish the connector to that assistant's directory. The name is what assistants display; the description is sent to the model as context rather than shown as a label. > **Is the configuration per environment?** > > Yes. Enablement, identity, and the tool selection are stored per project **and** environment. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.