openspec-knowledge-init

Generates and re-syncs the knowledge base taxonomy file from 1C configuration directories.

Updated Jun 26, 2026
One-click install
npx skills add https://github.com/pavelvdo/universal-xml-exchange2 --skill openspec-knowledge-init-pavelvdo
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: openspec-knowledge-init
Source: https://github.com/pavelvdo/universal-xml-exchange2/tree/main/.cursor/skills/openspec-knowledge-init
Command: npx skills add https://github.com/pavelvdo/universal-xml-exchange2 --skill openspec-knowledge-init-pavelvdo

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? When adopting a knowledge base in an existing 1C project, teams need an initial taxonomy that mirrors their actual configuration and extension layout, and they need to keep it in sync as configurations are added or removed without rerunning full project initialization. ## Core Features & Use Cases - Bootstrap taxonomy: Scans src/*/cf base configurations and src/*/cfe/* extensions to draft a _taxonomy.yaml with one domain per configuration, using deterministic ASCII transliteration for Cyrillic directory names. - Idempotent re-sync: Compares the draft against an existing openspec/knowledge/_taxonomy.yaml, shows a diff, and only adds missing domains or subdomains without deleting existing ones. - Confirmed writes: Presents a diff or summary and writes the taxonomy file only after explicit user confirmation, leaving _index.yaml and KB fact files untouched. - Use Case: After adding a new extension under src/ДО2 Подрядчик/cfe/, run the skill to propose a new domain like DO2_Podryadchik_cf with standard subdomains, review the diff, and confirm the update. ## Quick Start Ask the assistant to run /opsx:knowledge-init to bootstrap or re-sync the knowledge base taxonomy for the current project.

Frequently Asked Questions about openspec-knowledge-init

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I initialize a knowledge base taxonomy for an existing 1C project?

Run the /opsx:knowledge-init command, which scans your src/*/cf configurations and src/*/cfe/* extensions, drafts a taxonomy, shows a summary or diff, and writes openspec/knowledge/_taxonomy.yaml after your confirmation.

How are domain IDs generated from Cyrillic configuration folder names?

Domain IDs use deterministic transliteration: a leading ДО becomes DO, other Cyrillic letters are transliterated to ASCII, spaces and hyphens become underscores, and invalid characters are removed. For example, src/ДО2 Подрядчик/cf becomes DO2_Podryadchik_cf.

Does re-syncing the taxonomy delete existing subdomains or KB facts?

No. Re-sync is idempotent: it only adds missing domains and subdomains, never removes existing ones. The skill does not touch _index.yaml or any KB-*.md fact files.

Can the taxonomy be written without user confirmation?

No. The skill always shows a diff against the existing taxonomy or a summary for a new one, then waits for explicit user confirmation before writing openspec/knowledge/_taxonomy.yaml.

When should I use knowledge-init instead of full project initialization?

Use it when the project already exists and you only need the knowledge base taxonomy, either as a first bootstrap without /init-project or as a re-sync after adding or removing configurations and extensions.