BSL AI Toolkit

•
37
•

Enables AI-assisted development for 1C:Enterprise 8.3 (BSL) by providing a reusable set of rules, skills, MCP servers, and a byte-perfect metadata engine, ensuring AI operates on real code and documentation without hallucination.

The BSL AI Toolkit addresses the common challenge of AI models hallucinating details when developing for 1C:Enterprise 8.3, providing a robust and verifiable AI-first development environment. It offers a comprehensive, multi-agent framework comprising rules, skills, Model Context Protocol (MCP) servers, and a unique byte-perfect metadata engine. This toolkit is designed to be agent-agnostic, supporting various AI agents like Claude Code, Cursor, Copilot, and Gemini through a unified core and thin adapters, ensuring compatibility across different organizations and configurations. By operating directly on real 1C code, documentation, and diagnostics from tools like BSL Language Server, it ensures AI generates production-ready code and precisely manages complex metadata, effectively closing the 'last frontier' of AI-assisted 1C development.

主な機能

01Real-time code analysis and diagnostics using BSL Language Server (`bsl-ls`) and search across real configuration code (`onec-code`) for informed AI responses.
020 GitHub stars
03Byte-perfect metadata engine with round-trip verification for deterministic structural changes to 1C configurations (forms, layouts, attributes).
04Secure, read-only access to live 1C database data (`onec-data`) under investigated user permissions, including RLS and PII masking, for informed AI analysis.
05Multi-agent and provider-agnostic AI support, allowing integration with various LLMs (Claude Code, Cursor, Copilot, Gemini, etc.) via adaptable core and thin adapters.
06Agentic SWE capabilities through `onec-ops`, integrating with operational tools like Zabbix, Prometheus, and the 1C Technological Journal for performance and exploitation insights.

ユースケース

01Integrating AI into the full software development lifecycle for 1C, from requirements analysis to deployment and operational monitoring.
02Automating precise and verifiable modifications to 1C metadata, including forms, reports, and object attributes, with guaranteed byte-perfect results.
03Accelerating 1C:Enterprise 8.3 development and extension through AI-assisted coding, analysis, and quality assurance.