Finago launches an official MCP server for Procountor: AI that can work directly in your accounting system
Finago has today released an official MCP server for Procountor, making it possible to connect AI assistants directly to Procountor and let them work with your financial data, from answering questions to creating invoices. Greenstep works with Procountor every day through our accounting services, and our technology team of over 100 specialists builds AI and automation solutions grounded in real finance processes. That combination means we can help you connect the Procountor MCP server, set up safe permissions and approval practices, and design custom skills and agents that fit how your finance team actually works.
Finago has today released an official MCP server for Procountor, making it possible to connect AI assistants directly to Procountor and let them work with your financial data, from answering questions to creating invoices. The service is included in Procountor plans at no additional cost.
For finance teams, this is a significant step. Until now, using AI with accounting data has typically meant exporting reports and copying figures between tools. With the MCP server, you can simply ask an AI assistant to carry out tasks in Procountor, within the permissions you set.
What is MCP?
MCP, or Model Context Protocol, is an open standard that lets AI assistants connect to business systems in a controlled way. Think of it as a common plug: once a system offers an MCP server, compatible AI tools can read information from it and carry out actions in it, without custom integration work.
What the Procountor MCP server can do
According to Finago, the MCP server covers the functionality available through the Procountor API. In practice, that means an AI assistant can both retrieve information and take action in Procountor. For example, you can ask your assistant to create an invoice in plain language, and it will prepare it in Procountor for you.
It also works the other way: you can ask questions such as which invoices are overdue this month, or how this quarter’s costs compare to last year’s, and get answers based on your actual Procountor data.
The MCP server works with the common AI tools, such as Claude, ChatGPT and Microsoft Copilot,
Built-in AI skills and agents, and your own
Alongside the MCP server, Procountor comes with its own ready-made AI skills and agents for common financial tasks. Users can also create their own skills and agents in Procountor, tailored to how their company works.
This is where the real value lies for many companies. A ready-made agent handles the common tasks, while a custom agent can follow your own approval flows, reporting routines and ways of working.
Official release or community connector?
Community-built MCP connectors for Procountor have been available for some time. Finago’s release is the official version, included in Procountor plans and paired with Procountor’s own skills and agents. If you are already using a community connector, it is worth reviewing which option fits your security and support requirements.
What to get right before you connect
Because the AI can now create and change data in your accounting system, not just read it, access deserves the same care as any other user rights to your books. Before switching it on, we recommend deciding which users and AI tools get access, which actions the AI may carry out on its own and which need human approval, how AI-created invoices and entries are reviewed before they are sent or booked, and how the setup fits your company’s AI and data policies.
A good starting point for most companies is read access first, then gradually adding actions such as invoice creation once review practices are in place.
How Greenstep can help?
Greenstep works with Procountor every day through our accounting services, and our technology team of over 100 specialists builds AI and automation solutions grounded in real finance processes. That combination means we can help you connect the Procountor MCP server, set up safe permissions and approval practices, and design custom skills and agents that fit how your finance team actually works.