mizchi/skills

Agent skills by mizchi, distributed via APM

What it solves

This project provides a comprehensive library of "skills"—modular, standardized instructions and tools—that extend the capabilities of AI agents (such as Claude Code, Cursor, and Gemini CLI). It solves the problem of manually repeating complex technical workflows, audits, and best practices by providing pre-packaged, installable expertise across various domains like frontend development, cloud infrastructure, and formal methods.

How it works

Skills are distributed via the Agent Package Manager (APM) or installed using the skills CLI. Each skill follows the agentskills.io open standard, making them compatible across different agent runtimes. The repository is organized into categories (e.g., AWS, SQL, Node.js), where each directory contains a standalone skill. To ensure quality, the project includes waxa, a dedicated CLI tool for evaluating and iteratively tuning the prompts and performance of these skills using a structured evaluation framework.

Who it’s for

  • AI Agent Users: Developers who want to enhance their AI coding assistants with specialized knowledge in specific frameworks or cloud providers.
  • Agent Developers: Those building agentic workflows who need a standardized way to distribute and manage capabilities.
  • Technical Consultants: Professionals performing audits (e.g., frontend hygiene or security reviews) who can automate these checks via an agent.

Highlights

  • Broad Domain Coverage: Includes specialized skills for frontend triage, AWS OIDC patterns, SQL plan auditing, and MoonBit language practice.
  • Standardized Format: Adheres to the agentskills.io specification for cross-agent compatibility.
  • Evaluation Tooling: Features the waxa CLI for empirical prompt tuning and A/B testing of skill variants.
  • Easy Distribution: Integrated with APM for dependency management in apm.yml files.
  • VLM Integration: Includes vlmkit for vision-language model driven frontend audits and CSS auto-repair.

Related

  • Project
    hmmhmmhm/daiso-mcp

    A Model Context Protocol (MCP) server and CLI tool that connects AI agents to real-time South Korean retail inventory, convenience store data, cinema schedules, and fuel prices.

  • Project
    Wh1isper/mcp-email-server

    An MCP server that enables AI assistants to read, search, organize, and send emails using IMAP and SMTP.

  • Project
    docling-project/docling-parse

    Docling Parse is a fast, configurable PDF‑parsing library (Python API with a C++ core) that extracts text, words, lines, vector shapes and bitmap images together with their coordinates. It supports both sequential and multi‑threaded parsing, lets you choose what to compute or materialize per page, and includes rendering utilities. Ideal for AI workflows that need layout‑aware document data.

  • Project
    gameworkerkim/vibe-investing

    A comprehensive AI-driven investment ecosystem that combines multi-LLM agentic pipelines, quant regime classifiers, and corporate disclosure monitors to optimize financial decision-making.