Tool Reference

This page lists every tool the Generator Labs MCP server exposes, the scope it requires, and example prompts you can paste into your AI tool of choice. AI clients translate natural-language requests into the underlying tool calls automatically: you generally won't need to invoke them by name.

The list tools return up to 1000 items per call. When there are more, the response says it was truncated, and the AI tool can ask for the next page.

RBL Monitoring

rbl_list_hosts

List the RBL monitoring hosts on the account, with each host's current listing status, monitoring profile, contact groups, and tags. Supports filtering by status (active or paused) and by tag.

Scope: mcp:read

Example prompts:

• "Show me all the hosts I'm monitoring."
• "Which of my hosts tagged production are paused?"

rbl_get_host

Return the full record for a single monitored host: current listing status and listing details, monitoring profile, contact groups, and tags.

Scope: mcp:read

Example prompts:

• "Tell me more about the host mail.example.com."
• "What profile is 203.0.113.5 on, and is it listed?"

rbl_listings

Return every host currently listed on at least one RBL or URIBL data source, with the full per-source listing detail attached.

Scope: mcp:read

Example prompts:

• "What's currently listed across my account?"
• "Give me a quick summary of active listings."

rbl_get_listing_details

Per-host drill-down: which sources currently flag this host, what each one says, and how long the lookup took. Use after rbl_listings identifies hosts of interest.

Scope: mcp:read

Example prompts:

• "Why is mail.example.com flagged? Show me each source's detail."
• "Drill into the listing for the third host in that summary."

rbl_list_profiles

List the RBL monitoring profiles configured on the account, with their check frequency, timing window, and the data sources they use.

Scope: mcp:read

Example prompts:

• "What monitoring profiles do I have?"
• "Which profile uses the curated source list?"

rbl_listing_history

Return the listing history of a single host: one entry per check, with when it ran, whether the host was listed, on how many sources, and which sources flagged it. Looks back 30 days by default, up to 365, and returns up to 500 checks.

Scope: mcp:read

Example prompts:

• "Has mail.example.com been listed at all this month?"
• "Is the listing on 203.0.113.5 a recurring one?"

rbl_manual_check

Trigger an on-demand RBL check against a host. Accepts an existing host SID or a one-off literal IP / domain (not a subnet or range). Returns a check_id to poll. Counts against your daily manual-check allotment, and is billed per-check on Ultimate plans. An optional idempotency key makes a retry return the check already started instead of starting (and billing) another.

Scope: mcp:write

Example prompts:

• "Run an RBL check on mail.example.com right now."
• "Check if 203.0.113.5 is listed anywhere."

rbl_check_status

Poll the result of a manual check started via rbl_manual_check. Returns processing/complete state, listed/total counts, and (with include_details=true) the data sources that listed the host, or for an email-deliverability source, that failed it. Sources that didn't list the host are left out. Each source's details is its answer as plain text, and pass_fail appears only when it's true (an email-deliverability source).

Scope: mcp:read

Example prompts:

• "Is that check done yet?"
• "Show me the full results once it finishes."

rbl_list_sources

Resolve one or more RBL data source SIDs (the RB-prefixed identifiers that come back from manual checks and listing tools) to their friendly metadata: human name, DNS hostname, website URL, and type. Pass an optional sids array (up to 100) to look up specific sources, or omit it to list every source the account can see (the curated global pool plus your own custom sources).

Scope: mcp:read

Example prompts:

• "What's the friendly name for RB335b5143...?"
• "Map the source IDs from that last check back to readable names."
• "List every RBL source available on my account, grouped by type."

Certificate Monitoring

cert_list_monitors

List the certificate monitors on the account, with each monitor's hostname, protocol, state (good, error, expiring, or expired), days remaining, last check time, and current findings (see cert_get_monitor). Supports filtering by state. Set include_dismissed to also list each monitor's dismissed findings, after the others.

Scope: mcp:read

Example prompts:

• "Show me all my certificate monitors."
• "Which certs are in an error state right now?"

cert_get_monitor

Return the full record for a single certificate monitor: its state, expiry, fingerprint, protocol, and findings: every current finding, notices included, worst first. Each finding has its id (the finding key, such as CERT_HST02, explained in Certificate Findings), severity (error, warning or notice), message (its label) and dismissed. Findings you've dismissed are left out unless you set include_dismissed: then they're listed after the others, with dismissed true, so you can see what's dismissed and restore it with cert_dismiss_findings. A paused monitor has no findings. Every tool that returns findings uses this form.

Scope: mcp:read

Example prompts:

• "Tell me everything about the monitor for mail.example.com."
• "Why is my API endpoint's certificate monitor in error?"
• "Which findings have I dismissed on mail.example.com?"

cert_expiring_soon

List the certificates that expire within a given number of days (default 30, up to 365), and those that have already expired, each with its current findings in the same form as cert_get_monitor. Useful for building a renewal queue.

Scope: mcp:read

Example prompts:

• "Which of my certificates expire in the next 30 days?"
• "Have any of my certificates already expired?"

cert_recent_errors

List the active certificate monitors that currently have errors or warnings, such as an expired certificate, a certificate expiring soon, a name mismatch, a chain that isn't trusted, a failed connection, or a CAA problem. A monitor is listed when it has a warning or an error its profile alerts on: paused and archived monitors and monitors with only notices are left out, and a finding whose alert is turned off in the profile, or that you've dismissed, doesn't count. Each monitor's findings lists all its current findings, notices included, worst first, in the same form as cert_get_monitor; total_errors counts only the warnings and errors the profiles alert on.

Scope: mcp:read

Example prompts:

• "What cert monitors are in error state, and why?"
• "Which of my certificates have warnings I should look at?"

cert_get_event

Return the full result of a certificate check: the certificate's subject, common name, SANs, issuer, fingerprint, and expiry, and the result of each check on each IP address, with the finding key and severity of each problem. Pass a monitor SID for its latest check, or an event SID (CE-prefixed) for a specific earlier one.

Scope: mcp:read

Example prompts:

• "Show me the last check result for mail.example.com."
• "Which of the addresses behind www.example.com is serving the wrong certificate?"

cert_compliance_summary

An account-wide certificate compliance summary, matching the portal's Compliance Report: how many monitors are valid, expiring, expired, in error, or not checked yet. Scope it to one tag, or leave the tag out for a breakdown by tag. Give it an audit period in days (up to 365) to add the evidence for the period: total and failed checks, the percentage of checks passed, and the history of each monitor that had errors, for SOC 2, PCI DSS 4.0.1, or HIPAA audits.

Scope: mcp:read

Example prompts:

• "How compliant are my certificates right now?"
• "Were there any certificate issues in the last 90 days? I need it for our SOC 2 audit."
• "Which tag has the most certificate problems?"

cert_manual_check

Queue an on-demand check of one certificate monitor, for example right after you've renewed or replaced a certificate. The check is queued rather than run inline: read the monitor again with cert_get_monitor a few seconds later to see the result. A monitor can be checked this way at most once a minute and 10 times in 24 hours (100 per account). There is no charge beyond the monitor's normal daily charge. An optional idempotency key makes a retry return the check already queued instead of queuing another.

Scope: mcp:write

Example prompts:

• "I just renewed the certificate on mail.example.com. Check it now."
• "Re-check that monitor and tell me if the name mismatch is gone."

cert_dismiss_findings

Dismiss findings on certificate monitors where you've decided not to act on them, or restore them, with the same rules as the portal's Dismiss. Pass the finding keys in finding_ids, and optionally the monitor SIDs in monitors: without monitors, the findings are dismissed from every active monitor that has them now; with monitors, they're dismissed from each one named, whether or not it has them now. Set restore to true to restore instead: without monitors, that restores the findings on every monitor they're dismissed from. Only CERT_CAA01 (CAA Not Found), CERT_CAA02 (CAA Not Specific), CERT_LIF01 (Long Reg Period), CERT_HST01 (Matched CN, No SAN) and CERT_TRU03 (Root Sent) can be dismissed; any other key is refused. A dismissed finding is left out of the monitor's findings, cert_recent_errors and the alerts. The result lists the monitors changed, in hostname order, with the finding keys changed on each, and total counts them. Running it again changes nothing more.

Scope: mcp:write

Example prompts:

• "Dismiss all the CAA findings."
• "Restore the Root Sent finding on mail.example.com."

Notifications & Alerts

contacts_list

List notification contacts on the account: emails, SMS, Slack/Discord/Teams webhooks, PagerDuty/Opsgenie keys. Each row shows type, status, schedule, and which contact groups it belongs to. Sensitive credential fields are redacted.

Scope: mcp:read

Example prompts:

• "Who gets alerted when something fires?"
• "List all my Slack contacts."

contact_groups_list

List the contact groups on the account, with each group's member count, the number of hosts and certificate monitors that use it (total_monitors), and which group is the default.

Scope: mcp:read

Example prompts:

• "What contact groups do I have?"
• "Which group is my default, and who's in it?"

webhooks_list

List the webhook endpoints configured on the account, with their subscribed events and their delivery totals and failures over the last 90 days.

Scope: mcp:read

Example prompts:

• "What webhooks do I have set up?"
• "Are my webhooks healthy?"

Account & Health

account_summary

A high-level view of the account: company name, owner, plan, balance, the number of active and listed RBL hosts, and for certificate monitoring:

• active_monitors: the number of active monitors.
• monitors_with_findings: the number of active monitors with at least one current finding, of any severity.
• findings: the account's current certificate findings, like the By Finding view of the portal's Findings page, worst first. Each entry has its id (the finding key, such as CERT_CAA01), severity (error, warning or notice), message (its label) and monitors (the SIDs of the active monitors that have it, each of which shows the finding in its own findings). Certificates expiring soon are the CERT_EXP01 entry, and expired ones CERT_EXP02.

Findings you've dismissed are left out of both monitors_with_findings and findings.

Useful as a "give me the big picture" call.

Scope: mcp:read

Example prompts:

• "Give me an overview of my account."
• "What's my current balance and host count?"

mcp_health

Check that the MCP server can reach its own dependencies (its database and cache). Returns an ok flag, the server version, and the server time. Mostly useful for debugging a connection.

Scope: none

Example prompts:

• "Is the MCP server healthy?"

Server Prompts

In addition to tools, the server exposes three preset prompts: multi-step playbooks the AI tool can run end-to-end. Surface them by name in your AI tool's prompt picker:

Prompt	What it does	Arguments
cert_renewal_audit	Builds a renewal queue: every certificate expiring within the horizon, most urgent first, grouped by issuer and profile to show the ones that renew together.	horizon_days: how many days ahead to look (default 60)
incident_triage	Triages everything the account is dealing with now: hosts on RBLs, certificate errors, and certificates about to expire, most urgent first.	None
monthly_compliance_review	Walks through a monthly SOC 2 / PCI DSS / HIPAA compliance review: certificate health, RBL listings, and the audit evidence for the period.	days: the audit period in days (default 30)

Report Resources

With the mcp:reports scope, the server also lists the account's reports as MCP resources, which an AI tool can read and attach to a conversation: your 100 most recent RBL reports, and the latest check of each active or paused certificate monitor (up to 100).