.agents

Documents repository structure and maintenance conventions for a centralized AI agent skills collection.

Updated Jan 10, 2026
One-click install
npx skills add https://github.com/Rabbit-Ivan/Ivan-skills --skill agents-rabbit-ivan
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: .agents
Source: https://github.com/Rabbit-Ivan/Ivan-skills
Command: npx skills add https://github.com/Rabbit-Ivan/Ivan-skills --skill agents-rabbit-ivan

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Managing dozens of AI agent skills across multiple tools (Codex, Claude Code, and others) leads to version drift and duplicated copies. This Skill provides the repository-level entry point that explains how the Ivan-skills collection is organized, installed, and maintained from a single source of truth. ## Core Features & Use Cases - Repository Orientation: Explains the global management model where skills/ is the single source of truth synced to ~/.agents/skills. - Maintenance Conventions: Defines naming rules (kebab-case directories, mandatory SKILL.md), branch workflow (feature branches to main via PR), and which files are excluded from version control. - Use Case: When you need to add a new skill to the collection or understand how external upstream sources are tracked via .skill-lock.json, this entry point tells you exactly where things go and what conventions to follow. ## Quick Start Ask the agent to explain how this skills repository is organized and what steps are required to add a new skill under the skills directory.

Frequently Asked Questions about .agents

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

FAQPage Schema
How do I add a new skill to this repository?▼

Create a kebab-case directory under skills/ containing a SKILL.md file with name and description in its YAML frontmatter. Place optional scripts, references, and assets in their respective subdirectories, then submit via a feature branch and pull request to main.

How are AI agent skills installed from a GitHub repository?▼

Skills are installed using the npx skills add command pointing at the repository's skills directory, with flags for interactive selection, full installation, or single-skill installation. Installed skills land in ~/.agents/skills as the canonical copy.

Does this repository work with both Codex and Claude Code?▼

Yes. Codex reads skills directly from ~/.agents/skills, while Claude Code uses symlinks from ~/.claude/skills pointing to the same canonical copies, avoiding duplicated versions.

How are external upstream skill sources tracked?▼

External sources such as hylarucoder/hai-stack and larksuite/cli are recorded in .skill-lock.json with source URL, skill path, and content hash. This prevents version drift when the same skill exists in multiple locations.

What files are excluded from version control in this repository?▼

The .gitignore excludes .DS_Store, __pycache__ directories, compiled Python files, and the local backup directory skills_backup_20260214. These are local artifacts that should never be committed.