MCP (Model Context Protocol) is an open standard that lets AI assistants connect to external systems and pull live monitoring data. Generator Labs runs a hosted MCP server so tools like Claude, ChatGPT, Cursor, and GitHub Copilot can read your monitoring data and run on-demand checks. No integration code required.
What Is MCP?
MCP is to AI tools what USB-C is to peripherals. Before USB-C, every device needed its own cable. Before MCP, every AI tool needed custom integration code to talk to external systems. MCP standardizes the connection: one protocol, one server endpoint, any MCP-aware client.
The protocol defines how an AI tool discovers capabilities (tools), calls them, and gets results. It uses JSON-RPC 2.0 over HTTP and supports both API key and OAuth authentication.
MCP was created by Anthropic and is now supported across Claude Desktop, Claude.ai, ChatGPT, Cursor, VS Code with Copilot, and other tools. It is an open standard: the specification lives at modelcontextprotocol.io, and the reference implementations are on GitHub at github.com/modelcontextprotocol.
What the Generator Labs MCP Server Does
Each connected AI client automatically discovers the available tools. They span four areas:
For RBL monitoring: list hosts, check current listings, drill into per-source detail, run manual checks, and pull listing history. Ask which of your IPs are blacklisted right now and it will tell you, with per-source detail.
For certificate monitoring: list monitors, surface expiring certs, look up errors, and generate compliance audits. Ask for everything expiring in the next 30 days and you get the list; the compliance audit breaks your monitors down by tag.
For notifications: list contacts, contact groups, webhooks, and recent alerts, including whether deliveries succeeded.
For account: plan, balance, and host counts.
Example: Debugging a Blacklist Issue
You paste a hostname into Claude:
"Check if mail.example.com is listed on any RBL and show me the details."
Behind the scenes, Claude calls rbl_manual_check to trigger an on-demand scan, then polls rbl_check_status for results. Because you asked for the details, it sets include_details=true on that status call, which returns the per-source results, and calls rbl_list_sources to resolve internal IDs to human-readable names. What you see is a single answer listing each blacklist that flagged the host, what each source reported, and how long the lookups took.
No portal logins, no tabbing through screens, no manual copy-paste.
Example: Certificate Renewal Audit
"Audit my certificate compliance over the last 90 days."
Claude calls cert_compliance_summary, which returns your monitors bucketed by status (valid, expiring, expired, error, pending), an overall compliance percentage, a breakdown by tag, and an audit period covering total checks, error checks, and uptime. It is framed for SOC 2, PCI DSS, and HIPAA evidence. The AI summarizes where you stand and which tags are dragging the number down. To turn that into a work list, it follows up with cert_expiring_soon for the certs actually coming due.
How to Connect
The MCP endpoint is at https://api.generatorlabs.com/4.0/mcp. You can connect using either an API key or OAuth, depending on your tool.
Claude Desktop (API Key)
Add this to your claude_desktop_config.json:
{
"mcpServers": {
"generator-labs": {
"type": "http",
"url": "https://api.generatorlabs.com/4.0/mcp",
"headers": {
"Authorization": "Basic <your-base64-credentials>"
}
}
}
}
Combine your Account SID and Auth Token as AC1234...:abcdef... and Base64-encode the result.
Claude.ai / ChatGPT (OAuth)
Both support MCP via custom connectors. Add a connector with the URL above, click Connect, and complete the OAuth flow in your browser. You choose which API application the grant binds to and which scopes to approve.
What MCP Is Not
MCP is a conversational interface for querying and auditing, not a replacement for the REST API. You cannot create or modify monitors through MCP. For automation and scripting, the REST API and SDKs are the right tools. MCP is for asking your AI assistant a question about your infrastructure and getting an answer immediately.
See the AI monitoring page | MCP configuration | Full tool reference | Sign up