Using LookalikeGuard from an AI agent or assistant
If a user asks an assistant "is anyone impersonating my domain?" or "check for typosquats of example.com", the assistant can run a LookalikeGuard scan with one HTTP request and explain the results. Everything here uses the same public API that powers the website.
Quick facts
- What it does: finds registered lookalike domains (typos, look-alike characters, other extensions, added words like "login") and scores each 0 to 100 by risk, with the reasons.
- Data: public sources only: DNS, RDAP registration records, certificate transparency logs and each domain's public web page.
- Price: one-time scans are free. Monitoring: Solo $9/month, Team $29/month, Agency $79/month (yearly $90/$290/$790).
- Limits: monitoring and early warning only. No takedowns or legal advice. Results are automated and can contain mistakes.
Run a scan
GET https://lookalikeguard.com/api/v1/scan?domain=example.com
- No API key needed for one-time scans. There is a per-IP rate limit; over it you get HTTP 429.
- The request waits up to about 150 seconds. If the scan is still running you get HTTP 202 with an
id; pollGET /api/v1/results/{id}. - A finished result includes a
result_url(a shareable page at/r/<id>) you can give the user. - Scans of the same domain under an hour old are reused.
Machine-readable files
- /openapi.json: OpenAPI spec (also at /api/openapi.json; interactive docs)
- /llms.txt and /llms-full.txt: plain-text summary for language models
- /sitemap.xml
With the user's API key
Users on a paid plan can create an API key in Settings. With Authorization: Bearer <key> an agent can list monitored domains, fetch new findings since a time, and mark findings as mine, dismissed or watch. See API for developers.
Good manners for agents
- Only scan domains the user asked about.
- Do not sign up, start monitoring or pay on a user's behalf without their clear approval. Monitoring starts at /pricing.
- Tell the user scores are automated, and link the response guide if something looks risky.