AIVIS for developers

Authentication

Every request must be authenticated with an API key in the Authorization: Bearer aivis_live_... header. The key grants organization-level access (can read/trigger scans on any site in the organization).

Create your first API key from the dashboard

REST API v1

curl -X POST https://aivis.dev/api/v1/scan \
  -H "Authorization: Bearer aivis_live_..." \
  -H "Content-Type: application/json" \
  -d '{"domain": "example.com"}'

{
  "domain": "example.com",
  "score": 82,
  "letter": "B",
  "cached": false
}

curl https://aivis.dev/api/v1/report/example.com \
  -H "Authorization: Bearer aivis_live_..."

{
  "domain": "example.com",
  "score": 82,
  "letter": "B",
  "pagesCrawled": 12,
  "crawledAt": "2026-07-04T10:00:00.000Z",
  "sector": "saas",
  "checks": [
    { "id": "robots-ai-bots", "label": "AI bot access", "status": "pass", "score": 100, "evidence": ["..."], "fixHint": null }
  ]
}

<a href="https://aivis.dev/report/example.com">
  <img src="https://aivis.dev/api/v1/badge/example.com" alt="AIVIS AI Visibility Score" />
</a>

Every endpoint returns 401 for a missing/invalid key and 403 if the organization isn't on a Business or Agency plan.

MCP server

AIVIS is also available as an MCP (Model Context Protocol) server: a compatible AI assistant can run scans and read reports directly, without going through a browser. Endpoint: POST https://aivis.dev/api/mcp, JSON-RPC 2.0, same Bearer authentication as the REST API.

• scan_domain — trigger (or fetch from cache) a full scan
• get_report — read the latest saved report without triggering a new scan
• check_ai_access — quick check (robots.txt only) of which AI bots are allowed or blocked

Example configuration for an MCP client that supports custom headers (e.g. Claude Desktop):

{
  "mcpServers": {
    "aivis": {
      "url": "https://aivis.dev/api/mcp",
      "headers": {
        "Authorization": "Bearer aivis_live_..."
      }
    }
  }
}