Guardio

How many online services are linked to the account and how many are hidden.

report_posture

Check one URL or hostname for phishing or scams. For a message or email, extract the relevant URLs locally and call this tool separately for each URL. Do not send message bodies or conversation history. Returns dangerous, suspicious, or no_known_threat. no_known_threat means Guardio knows of no threat, not that the site is guaranteed safe. If trusted_by_user is true, the result reflects the user's trust setting. Guardio records each check on the account, and a dangerous result is added to the account's activity.

check_link

Check whether a phone number appeared in known data breaches. Use international format with the country code, for example +14155550123. Returns how many breaches, what kinds of data were exposed and the breach names and years; never the leaked data itself.

check_phone_leaks

Check whether an email address appeared in known data breaches. Returns how many breaches, what kinds of data were exposed and the breach names and years; never the leaked data itself.

check_email_leaks

The connected Guardio account: email, plan name and device allowance, subscription status, role on the plan and whether it is a trial. Takes no arguments. Call it first when unsure what the account can see.

whoami

Per-feature setup state (on, partial, off) rolled up across the account.

report_features

Record the user's feedback about these Guardio tools: a missing feature, a bug or a wrong answer. Call it only when the user asks you to leave feedback, or agrees after you offer. It stores one note in this Guardio service, for the Guardio team to read. It sends no email, message or ticket, and nothing is published or shared outside Guardio. Describe what happened and what you expected in a few plain sentences, and name the tool if there is one. Leave out passwords, card numbers, message bodies and other personal data. Record one note per problem, not one per retry. Guardio does not reply, so do not promise the user an answer.

send_feedback

Breaches that exposed the account's monitored emails or phone numbers.

report_recent_leaks

Emails and pages scanned, flagged, blocked and warned in the last 7 and 30 days.

report_scans_counters

The account's security recommendations, most urgent first: what each one is, why it matters and the action to take. Pending ones need attention; resolved and dismissed are history.

report_recommendation_content