Agent access — API & MCP
The audit isn't just for people. An AI agent can run it directly — over a documented REST endpoint or the Model Context Protocol.
MCP server
The MCP endpoint is a stateless Model Context Protocol server (Streamable HTTP transport). Point an MCP-compatible client at:
https://futureproofwebsite.com/mcp
It exposes two tools:
audit_site(url)— score a website against the AI Readiness Index.get_ai_readiness_index()— return the framework the score is based on.
Example handshake and tool call (JSON-RPC 2.0 over HTTP POST):
curl -s https://futureproofwebsite.com/mcp \
-H 'Content-Type: application/json' \
-H 'Accept: application/json, text/event-stream' \
-d '{"jsonrpc":"2.0","id":1,"method":"initialize",
"params":{"protocolVersion":"2025-06-18","capabilities":{},
"clientInfo":{"name":"demo","version":"1.0"}}}'
curl -s https://futureproofwebsite.com/mcp \
-H 'Content-Type: application/json' \
-H 'Accept: application/json, text/event-stream' \
-d '{"jsonrpc":"2.0","id":2,"method":"tools/call",
"params":{"name":"audit_site","arguments":{"url":"https://example.com"}}}'
REST endpoint
For non-MCP clients, the same audit is available as JSON:
GET https://futureproofwebsite.com/api/audit?url=https://example.com
Returns the full scored result — per-pillar scores, each criterion's status and detail, and the criteria flagged for manual review.
Notes
- The scanner refuses private/loopback addresses, rate-limited and abuse-resistant by design (criterion AS2, agentic defensibility).
- The automated tier covers the structural criteria; Core Web Vitals, accessibility conformance, provenance and deeper agent-surface review are flagged for the full report.
- See the framework behind the score: the AI Readiness Index.