Back to directory

Claude Code Skills for 1C:Enterprise

A set of 116 Claude Code skills, rules and commands that let an agent assemble and edit 1C:Enterprise metadata objects, forms and extensions from compact JSON and validate them against the rest of a configuration.

Claude Code Skills for 1C:Enterprise

This project gives Claude Code the ability to edit 1C:Enterprise source in the platform's own XML export format rather than by searching plain text files. Skills let the agent assemble or parse metadata objects, forms, roles, layout schemes and extensions, while separate rules encode BSL coding standards, anti-patterns and requirements from the 1C:EDT development environment and the BSP standard subsystems library.

Fourteen validators check the generated XML for an object, form, role, layout scheme or extension, and a configuration index lets them cross-check an object against the rest of the configuration, such as reference types, form data paths, subsystem membership and role rights. None of this requires the 1C platform, EDT or a running database; the validators check names and structure only, not BSL expression types or semantics.

The set also includes reference catalogs of BSP library methods and usage patterns, and of 1C:ZUP payroll and HR functions, so the agent looks up real method signatures instead of guessing them. It is written in Python and MIT licensed.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.