MCP-сервер для 1С

MCP-1C is a solution designed for integrating AI assistants with 1C databases, offering both a gateway for company-wide deployment and a personal server for developers. The platform addresses secure access to 1C systems, enabling employees and developers to interact with 1C data through AI assistants while maintaining strict controls over permissions, privacy, and auditability.

For organizations, MCP-1C provides a self-hosted gateway that operates within the company's infrastructure. This gateway mediates all assistant requests to 1C databases, applying role-based access controls, masking personally identifiable information (PII), enforcing query quotas per database, and maintaining a comprehensive audit log of all assistant activities. It supports a safe mode that restricts AI assistants to read-only access on production databases, blocking write operations to protect critical data. The gateway also features real-time notifications for blocked operations or quota breaches, and exposes operational metrics such as load, errors, and assistant activity to engineering teams for monitoring and observability.

For developers, MCP-1C offers a personal server that can be connected directly to an IDE or AI assistant, allowing work with 1C databases from within the development environment. The server is easy to set up, requiring only a single executable file and a simple configuration. The open edition of the server is available as open source and can be used for free, with its source code accessible for review and contributions. Paid editions introduce additional features: the extended edition includes a query optimizer, BSL code linter, and static analysis tools, while the professional edition adds bulk analysis, security auditing, and semantic search capabilities.

MCP-1C is suited for both enterprise IT departments seeking secure, auditable AI integration with 1C, and individual developers looking for enhanced tooling within their workflow. Its self-hosted architecture ensures that access policies and data privacy remain under organizational control, while its tiered editions allow users to select the level of functionality appropriate to their needs.