Skills

A collection of CLI tool skill definitions for AI assistants.

Available Skills

Skill	Description
ast-grep	Structural code search with ast-grep using AST-aware patterns
google-workspace	Google Workspace MCP integration through maco - Gmail, Calendar, Contacts, and Drive
librarian	Maintain a local cache of remote Git repositories for code research and exploration
icloud-cli	Manage iCloud calendars, events, and email via the CLI
matchlock	Run AI agents in ephemeral micro-VMs with VM-level isolation, network allowlisting, and secret injection - CLI, Go SDK, and Python SDK
schedule	Manage scheduled, background agentic tasks
strix-halo-llm	Run, benchmark, and serve local GGUF models on AMD Strix Halo using kyuz0 toolboxes, memory estimation, and reproducible podman commands
tmux	Run interactive CLIs and long-running tasks in isolated tmux sessions - manage session lifecycle, scrape pane output, and poll for patterns
uv	Astral uv usage for uv run, PEP 723 inline script dependencies, dependency compilation/locking, and uv_build backend setup
waitrose-cli	Manage Waitrose grocery shopping - trolley, product search, delivery slots, and orders

Available Recipes

Recipe	Description
code/architect	Analyzes codebase patterns and designs architectural solutions with implementation blueprints
code/reviewer	Performs a comprehensive code review of changes with focus on quality, security, and best practices
ralph/init	Generate a PRD (Product Requirements Document) based on discussion and design docs for autonomous development
ralph/iterate	Iteratively work through a feature list, implementing one feature at a time with progress tracking

Available Extensions

Extension	Tools	Description
code-search	code_search	Agentic codebase search for complex, multi-step code discovery tasks
look-at	look_at	Targeted analysis of local files, including PDFs, images, audio, video, and documents
nano-banana	nano_banana	Generate images with Gemini Nano Banana and save them under ~/.cache/nano-banana
subagent	subagent	Delegate self-contained tasks to an independent Kodelet session

Structure

Each skill is contained in its own directory with a SKILL.md file that provides:

• Metadata (name, description, trigger conditions)
• Prerequisites and setup instructions
• Command reference and usage examples
• Common workflows and troubleshooting tips

Extensions live under extensions/ as executable kodelet-extension-* Python SDK scripts with inline uv dependency metadata.

Installation

# Install this plugin repository globally
kodelet plugin add jingkaihe/skills -g

# Install locally for the current repo
kodelet plugin add jingkaihe/skills

Usage

These skill definitions are designed to be loaded by AI assistants to enable interaction with external CLI tools. Each SKILL.md follows a standard format with YAML frontmatter for metadata.