Cin7 Omni MCP server for your AI

Two values from Cin7 Omni: an API Username and an API Key. You paste both into the connect form and the connection authenticates server-side over HTTPS. There is no ongoing re-authentication to keep alive.

75 operations across 14 API modules, split 43 read and 32 write. 15 load into your AI client by default and the rest stay searchable, so the tool list does not swamp the client. Stock itself is read-only: list_stock returns a snapshot per product option and branch rather than letting you set a number, and delete_payment is the only delete among the tools listed on this page.

Cin7 Omni has no separate StockTake resource, which is the thing that trips most people up. Stocktakes are reconciled as stock adjustments, so the agent reads the current position with list_stock and list_branches, then posts the counted difference with create_adjustments. If you have approval-required-for-writes on, that adjustment waits for you before it lands.

Both live in the same contacts resource. get_contact returns a record with a type field that says whether it is a customer or a supplier, list_contacts covers both, and create_contacts creates either one. Ask for suppliers and the agent filters on that field instead of calling a separate endpoint.

Yes, manufacturing is one of the bigger modules here. list_bom_masters and get_bom_master read the v1 model while list_bom_masters_v2 and get_bom_master_v2 return the v2 read model, so point the agent at the one whose fields you expect; writes go through create_bom_masters and update_bom_masters. Production jobs are full read and write with list_production_jobs, get_production_job, create_production_jobs and update_production_jobs.

Every trigger on this connection is polled, so a sales order stage change, a dispatch, a void, a payment recorded, a payout received or a credit note issued shows up on the next poll rather than as an instant push. Coverage spans the order lifecycle, quotes created and accepted, purchase orders raised and received, new customers or suppliers, stock adjustments and branch transfers. Build workflows that tolerate a short lag rather than expecting sub-second reaction.

No. PopMCP hosts it and gives you a Streamable HTTP MCP URL. There is nothing to deploy, no callback URL to register, and no infrastructure to keep running or patch.

Any client that supports a remote Streamable HTTP MCP server — Claude Desktop, Claude Web, ChatGPT custom connectors, Codex and the Codex CLI, OpenCode and Z.ai's ZCode, as well as editors such as Cursor, Windsurf and Zed. Your client signs in to PopMCP with OAuth; the provider credential never leaves the server.

Loading 75 tool definitions into every conversation burns context before the work starts. PopMCP loads a curated 15-tool set covering the everyday workflows and keeps the rest searchable, so your AI pulls a specialist tool only when a task needs it. You can switch a connection to the full catalog at any time.

A tool is something your AI calls — there are 75 here. A trigger is something PopMCP watches for you: 19 Cin7 Omni events that can start work on the PopMCP side. Triggers never enter your AI client's context, and none of them ask you to paste a callback URL into Cin7 Omni.