Getting Started — Integrate with AX Audit

There are none.AX Audit’s own API and MCP server need no signup, no API key, and no account creation. Every endpoint under /api/v1/public/* is open, and the MCP server at https://ainative.ax/mcp requires no authentication for either of its tools. If you can send an HTTP request or connect an MCP client, you can integrate right now.

This is a deliberate choice, not an oversight — it is itself one of the things AX Audit checks for in other sites, under the AR-PROV (Agent Provisioning Readiness) category. See the AR-PROV guide for why zero-friction provisioning matters for the agentic web generally.

There are two independent integration paths, and you can use either one or both. Pick the REST API for CI/CD pipelines, scripts, and backend services. Pick the MCP server for an agent or IDE assistant (Claude Code, Claude Desktop, Cursor, or any MCP-compatible client) that should be able to run scans as part of a conversation.

None, for either path.The REST API’s public endpoints and the MCP server’s tools are both open — no bearer token, no OAuth flow, no API key header. Rate limits apply only to abusive traffic; there is no published per-key quota today. If you’re integrating a high-volume CI/CD or monitoring use case, contact us directly.

Base URL: https://api.ainative.studio. All endpoints live under /api/v1/public/ and use application/json for both requests and responses. Set a client timeout of at least 60 seconds for a quick scan and 90 seconds for a full report — a cold target site can take that long to render and probe.

The two endpoints most teams start with:

POST /api/v1/public/ax-audit
# Quick scan — score 0-100 + per-check breakdown. Not persisted.
{ "url": "https://example.com" }

POST /api/v1/public/ax-audit/full
# Full report — adds an agent-simulation task + industry benchmark.
# Persisted; returns a report_id you can fetch again later.
{ "url": "https://example.com", "industry": "saas" }

Quick scan, in curl:

curl -X POST https://api.ainative.studio/api/v1/public/ax-audit \
  -H "Content-Type: application/json" \
  -d '{"url": "https://example.com"}'

Full field reference — every endpoint, request/response shape, error codes, and code examples in curl, Python, and JavaScript — is on the API documentation page.

AX Audit runs a real, hosted MCP server at https://ainative.ax/mcp over Streamable HTTP. There is nothing to npx install or run locally — it is a remote server your MCP client connects to directly, the same way it would connect to any other remote MCP server. The server publishes its own card at /.well-known/mcp.json for agents that discover tools that way.

Add it in Claude Code with the CLI:

claude mcp add --transport http ax-audit https://ainative.ax/mcp

Or add it directly to a client config (Claude Desktop and others that read an mcpServers block):

{
  "mcpServers": {
    "ax-audit": {
      "url": "https://ainative.ax/mcp"
    }
  }
}

Once connected, two tools are available, neither requiring authentication:

For the deeper mechanics of MCP discovery (server cards, .well-known/mcp.json, and how agents find MCP servers generally), see the MCP Discoverability guide.

This page covers what you need to start integrating. For depth on any one piece:

• /api — full REST API reference: every endpoint, request/response shapes, timeouts, rate limits, error codes.
• /learn/mcp-discovery — how MCP discovery works generally, and the server-card format.
• /methodology — how the AX Audit score itself is calculated, and what it can and can’t tell you.
• /learn/ar-prov — the AR-PROV standard for zero-friction agent provisioning, which this site practices.