> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tuteliq.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# MCP Server for Content Moderation and Child Safety

> Connect Claude, Cursor or any MCP client to Tuteliq over OAuth. Around 80 tools for grooming detection, content moderation, incident review and compliance, with no API key to configure.

Tuteliq runs a hosted [Model Context Protocol](https://modelcontextprotocol.io) server at `https://api.tuteliq.ai/mcp`. Point an MCP client at that URL, sign in through the browser, and an AI assistant gains around 80 tools spanning detection, moderation, verification and compliance.

Nothing to install, and no API key to paste into a config file.

## Connect

<Steps>
  <Step title="Add the server">
    **Claude Desktop:** **Settings > Connectors**, **Add custom connector**, name it **Tuteliq**, URL `https://api.tuteliq.ai/mcp`.

    **Claude Code, Cursor, Windsurf and other clients supporting remote servers:**

    ```json theme={"dark"}
    {
      "mcpServers": {
        "tuteliq": {
          "type": "http",
          "url": "https://api.tuteliq.ai/mcp"
        }
      }
    }
    ```
  </Step>

  <Step title="Authenticate">
    Connect. Your browser opens a Tuteliq sign-in page; approve, and the client stores the token. In Claude Code, run `/mcp` if the flow does not open on its own.
  </Step>
</Steps>

Headless environment, or a client without remote server support? See [authentication](/mcp-server/authentication) for static tokens and stdio.

## What you can build

<CardGroup cols={3}>
  <Card title="An agent that works your queue" icon="shield-halved" href="/mcp/moderation">
    Size the backlog, triage by severity, read what matters, record decisions into the same audit trail a human moderator's land in.
  </Card>

  <Card title="A safety layer in an assistant" icon="comments" href="/mcp/tools">
    Screen conversations for grooming, bullying, self-harm or fraud before they reach a human.
  </Card>

  <Card title="A compliance assistant" icon="file-shield" href="/mcp/tools#governance-and-compliance">
    Pull incident volumes and dispositions for a DSA or KOSA pack, fetch signed audit receipts, answer GDPR requests.
  </Card>
</CardGroup>

## What it covers

| Area                       | Examples                                                                 |
| -------------------------- | ------------------------------------------------------------------------ |
| **Detection**              | Grooming, bullying, unsafe content, emotions, multi-endpoint analysis    |
| **Fraud and exploitation** | Romance scams, mule recruitment, coercive control, radicalisation        |
| **Media**                  | Image, voice, video and PDF analysis, plus synthetic-media detection     |
| **Moderation**             | Queue triage, incident detail, trends, moderator decisions               |
| **Governance**             | Signed audit receipts, GDPR endpoints, consent ledger, breach register   |
| **Operations**             | Policy rules, detection settings, webhooks, usage, verification sessions |

See the [full tool reference](/mcp-server/tools) for all \~80.

<Note>
  Detection quality is identical to the REST API. The MCP server is a transport, not a different model: both are backed by the same tool definitions.
</Note>

## Two things worth knowing early

**Agents cannot resolve incidents.** An MCP token identifies an account, never an individual, so an agent is limited to `confirm` plus read-only work. Escalation, resolution and reopening stay with a human moderator. See [moderation](/mcp-server/moderation#what-an-agent-may-decide).

**No user content is stored.** Messages, images, audio, video and documents are processed in memory and never persisted, raw or derived. What exists is the abstracted detection record. Customers running their own moderation stack can switch off even that, via [incident logging control](/incident-logging).

## Next steps

<CardGroup cols={2}>
  <Card title="Authentication" icon="key" href="/mcp/authentication">
    OAuth 2.1, token lifetime, refresh rotation, static tokens and stdio.
  </Card>

  <Card title="Moderation workflows" icon="list-check" href="/mcp/moderation">
    Working a queue with an agent, and what it is allowed to decide.
  </Card>

  <Card title="Tool reference" icon="wrench" href="/mcp/tools">
    All \~80 tools by category, and everything that changes or deletes data.
  </Card>

  <Card title="Credits" icon="coins" href="/credits">
    What each tool costs. Text is 1 to 3, an image 7, a voice clip 21, a video 95.
  </Card>
</CardGroup>