Command Claude Code
Build release artifacts from committed sources — .cf for the main configuration and .cfe per extension, with optional version bump, changelog and git tag.
Rules\skills\subagens for vibecoding in 1C(bsl)
This repository also configures its own agents. See what ai_rules_1c tells them →
Command Claude Code
Build release artifacts from committed sources — .cf for the main configuration and .cfe per extension, with optional version bump, changelog and git tag.
Command Claude Code
Toggle the caveman communication style — set CAVEMAN (on|auto|off) in .dev.env for the whole project; on|off|auto persist, lite|full|ultra switch the session level.
Command Claude Code
Check a 1C XML configuration dump for duplicate UUIDs, and optionally regenerate them.
Command Claude Code
Check availability of 1C MCP servers and install/start the missing ones.
Command Claude Code
Read-only check for available updates of the 1C MCP server images and of the 1c-rules ruleset, without pulling or installing anything.
Command Claude Code
Load the configuration into the test infobase from .dev.env and check the result in the 1C interface — QA MCP first, the web client as fallback.
Command Claude Code
Diagnose whether 1c-rules is installed, connected, configured, and usable by the current agent.
Command Claude Code
Toggle orchestrator economy mode via ORCHESTRATION in .dev.env; on enable, offers to configure per-tier subagent models and to set up the rtk output-compression proxy.
Command Claude Code
Aggregate accumulated rule-friction signals into user-approved behavior rules in LLM-RULES.md (ruleset self-improvement loop).
Command Claude Code
Extract configuration objects from infobase to files for editing.
Command Claude Code
Initialize a 1C project — test infobase from an existing IB / .cf / .dt, full source dump (cf + all cfe) into a git repository, first commit.
Command Claude Code
Install agent-browser (token-efficient headless browser CLI + MCP) for 1C web UI testing.
Command Claude Code
Install Cognee MCP as the preferred persistent memory write provider and register it in the active AI client.
Command Claude Code
Install DitriX EDT-MCP into a local 1C:EDT installation and connect the active AI client.
Command Claude Code
Install Windows-MCP (desktop UI automation MCP) as a last-resort alternative to home-grown screenshot/OCR.
Command Claude Code
Download the 1C MCP server distribution from vibecoding1c.ru and install all servers from it.
Command Claude Code
Detect and optionally install the supported 1C development tools through one guided menu.
Command Claude Code
Compatibility alias for /sdlc profile names; UI testing is controlled separately by /uitests.
Command Claude Code
Dump the configuration from the infobase defined in .dev.env into the current repository files.
Command Claude Code
Rebuild the test infobase to the effective snapshot — optional DT data baseline, then the current cf + cfe sources from git.
Command Claude Code
Adapt the ruleset to the model that is actually running — normalize a free-form model name to a profile slug and write AGENTMODEL (opus5|sonnet5|fable5|gpt56|gpt6) into .dev.env.
Command Claude Code
Send a problem report about a 1C MCP server or about the 1c-rules ruleset to the support service (Yandex Cloud Function -> Yandex Database).
Command Claude Code
Show the status of your support tickets about MCP servers and the 1c-rules ruleset, and close a ticket you no longer need.
Command Claude Code
Closed deploy → test → fix → redeploy loop against the test infobase, until the scenarios pass or the iteration budget is exhausted.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: