# Assistants

Chat assistants and local chat apps that can call our MCP server. You add localscreenshot once as a custom connector, then ask for a screenshot in the conversation.

Most of them have a field for a header or a token. That is where your key goes:

**MCP server**

```text
Address    https://api.localscreenshot.com/mcp
Transport  HTTP
Header     Authorization: Bearer $LOCALSCREENSHOT_API_KEY
```

A few have no place for a key and connect by sign-in instead: the assistant opens a localscreenshot window, you sign in and allow it. Those are marked below, and stay listed as not ready until that way of connecting is open.

The [dashboard](https://localscreenshot.com/dashboard) shows the same steps with your own key already in place.

### Claude

In Claude (web or desktop): Customize, Connectors, then add a custom connector. Write your key in place of `$LOCALSCREENSHOT_API_KEY`.

```text
1. In Claude, open Customize > Connectors
2. Click "+ Add", then "Add custom connector"
3. Name: localscreenshot
4. URL: https://api.localscreenshot.com/mcp
5. Under Request headers, add
   Authorization: Bearer $LOCALSCREENSHOT_API_KEY
6. Save, then ask Claude for a screenshot
```

From its documentation, not tried by us yet. Source: [support.claude.com](https://support.claude.com/en/articles/11175166-getting-started-with-custom-connectors-using-remote-mcp).

### Le Chat

In Le Chat: Connectors, Add Connector, Custom MCP Connector. Write your key in place of `$LOCALSCREENSHOT_API_KEY`.

```text
1. Open the Connectors page and click "+ Add Connector"
2. Switch to the "Custom MCP Connector" tab
3. Connector name: localscreenshot
4. Server URL: https://api.localscreenshot.com/mcp
5. Authentication: HTTP Bearer Token
   $LOCALSCREENSHOT_API_KEY
```

From its documentation, not tried by us yet. Source: [docs.mistral.ai](https://docs.mistral.ai/le-chat/knowledge-integrations/connectors/mcp-connectors).

### Copilot Studio

In your agent: Tools, Add a tool, New tool, Model Context Protocol. Write your key in place of `$LOCALSCREENSHOT_API_KEY`.

```text
1. On your agent's Tools page: Add a tool > New tool > Model Context Protocol
2. Server name: localscreenshot
3. Server description: Takes screenshots of public web pages
4. Server URL: https://api.localscreenshot.com/mcp
5. Authentication: API key, Type: Header, Header name: Authorization
6. Create, then enter this value as the API key
   Bearer $LOCALSCREENSHOT_API_KEY
```

From its documentation, not tried by us yet. Source: [learn.microsoft.com](https://learn.microsoft.com/en-us/microsoft-copilot-studio/mcp-add-existing-server-to-agent).

### LM Studio

In LM Studio: Program tab, Install, Edit mcp.json. Write your key in place of the variable. This tool does not read environment variables in its settings.

```json
{
  "mcpServers": {
    "localscreenshot": {
      "url": "https://api.localscreenshot.com/mcp",
      "headers": { "Authorization": "Bearer $LOCALSCREENSHOT_API_KEY" }
    }
  }
}
```

From its documentation, not tried by us yet. Source: [lmstudio.ai](https://lmstudio.ai/docs/app/mcp).

### Open WebUI

In Open WebUI: Settings, Admin, Integrations, External Tool Servers. Write your key in place of `$LOCALSCREENSHOT_API_KEY`.

```text
1. Open Settings > Admin > Integrations
2. Under External Tool Servers, click "+ Add Connection"
3. Type: MCP (Streamable HTTP)
4. Server URL: https://api.localscreenshot.com/mcp
5. Auth: Bearer
   $LOCALSCREENSHOT_API_KEY
6. Save
```

From its documentation, not tried by us yet. Source: [docs.openwebui.com](https://docs.openwebui.com/features/extensibility/mcp/).

### LibreChat

Add to librechat.yaml. LibreChat reads the key from its environment.

```yaml
mcpServers:
  localscreenshot:
    type: streamable-http
    url: https://api.localscreenshot.com/mcp
    headers:
      Authorization: 'Bearer ${LOCALSCREENSHOT_API_KEY}'
```

From its documentation, not tried by us yet. Source: [www.librechat.ai](https://www.librechat.ai/docs/configuration/librechat_yaml/object_structure/mcp_servers).

### Not ready yet

We show a setup only once we have read it in the tool's own documentation. Not yet for: ChatGPT, Perplexity.

## Then ask

> "Take a screenshot of wikipedia.org on an iPhone."

A connector you add is yours: the screenshots it takes use your credits and appear in your dashboard, like any other. What an assistant can ask for is on [Ask your AI (MCP)](https://docs.localscreenshot.com/agent).