PBuilt with MG Page BuilderThis entire website — header, pages and footer — is designed with our own WHMCS page builder.See how it works
Connect Claude, Codex or Cursor to your WHMCS — with guardrails.
A Model Context Protocol server that runs inside WHMCS. 170 tools, per-token permissions, risk-based approvals and a full audit log, so AI agents can work on clients, billing, services and tickets without getting the keys to everything.
Point any MCP client at your WHMCS endpoint. The agent picks the right tools, reads live data and prepares changes — while the module decides what it is allowed to do.
Search invoicesBilling · read✓ 38 msClient overviewClients · read✓ 3 clientsThree clients have unpaid invoices above $100 in September:
Suspend serviceServices · write⏳ approval requiredThis is a high-risk action, so I created draft #218 instead of running it. An administrator can approve it from the Approvals tab.
Each token has its own mode, permissions, IP allowlist and expiry. The global mode can never expand a token beyond what you granted.
Choose how mutations are handled. Every change can go through a draft that is validated, approved and only then executed.
Suspend serviceService #4412highApproveCancel
#217Update pricingProduct #12highApproveCancel
#216Apply creditClient #88 · $20mediumExecuted
#215Reply to ticketTicket #5531lowExecuted
#214Close clientClient #61criticalCancelled
Client overview203.0.113.24ok
22:40:52Claude DesktopSearch invoices203.0.113.24ok
22:38:15Cursor botSuspend service198.51.100.7draft
22:31:40Cursor botUpdate client198.51.100.7denied: permission
22:29:03unknownList clients192.0.2.99denied: IP
22:12:44Claude DesktopReply to ticket203.0.113.24ok
Every action is logged with its token, IP and outcome — including the ones that were refused.
From read-only reporting to day-to-day operations. Each area has its own read and write permissions, so every agent only reaches what you allow.
170 tools, scoped tokens, approvals and a full audit trail.
Buy NowPowered by WHMCompleteSolution