Question 1

Which IP addresses can I look up?

Accepted Answer

Use GET /ip-lookup or GET /fraud-check with the optional ip query parameter containing a public IPv4 or IPv6 address, for example 8.8.8.8. Private, loopback and reserved addresses are rejected. If ip is omitted, the service tries to detect the client IP from trusted gateway forwarding information. When integrating behind a proxy, pass the public IP you intend to evaluate explicitly.

Question 2

What does the fraud score mean, and is VPN detection guaranteed?

Accepted Answer

The fraud_score ranges from 0 to 100 and combines external IP risk signals with hosting-network heuristics and available bot signals. is_high_risk is true at 70 or above. Scores are indicators for review, not proof of fraud, and should not be the sole reason to block someone. VPN, proxy and Tor signals can be null when intelligence is unavailable. Geolocation and risk data are supplied by ipquery.io, with Cloudflare edge metadata when applicable; external results may be cached for up to six hours.

Question 3

How can an AI agent look up IP risk through MCP?

Accepted Answer

Connect to https://mcp.jojapi.com/fraudguard-and-ip-risk-intelligence-api using an HTTP-compatible MCP client. Configure Authorization: Bearer YOUR_JOJ_API_KEY as a connection header and select a plan. Pass the public IP explicitly. Use ip-lookup for geolocation and ASN context, or fraud-check for network-risk signals. Agent calls use the same subscription and quotas as normal requests. Preserve null or unknown signals and treat the score as decision support, not proof of fraud.

IP Geolocation and VPN Proxy Risk Detection

Frequently Asked Questions