This repository provides an official, unified collection of skills for the GLM family of AI models, designed to enhance agent architectures like Claude Code, OpenCode, and AutoClaw. It consolidates functionalities previously scattered across individual model repositories, offering developers a centralized resource for advanced AI capabilities.

How It Works

The project acts as a curated library of specialized "skills" that extend the functionality of GLM models. These skills leverage various AI techniques, including multimodal processing, optical character recognition (OCR), and generative AI, to perform complex tasks. The core approach is to modularize these capabilities, allowing integration into diverse AI agent frameworks and simplifying access to advanced features like document analysis, image generation, and application development.

Quick Start & Requirements

Installation is recommended via Clawhub:

• Single skill: npx clawhub@latest install glmocr
• Multiple skills: npx clawhub@latest install glmocr glmocr-table glmv-caption glm-image-gen

Alternatively, clone the repository (git clone https://github.com/zai-org/skills.git) and follow individual SKILL.md instructions. Most skills require a ZHIPU_API_KEY environment variable, obtainable from bigmodel.cn.

Highlighted Details

• GLM-V (Multimodal): Includes skills for image/video captioning, document-based writing, object localization, PDF-to-presentation/web conversion, PRD-to-app generation, AI art prompt generation, resume screening, stock analysis, and web replication.
• GLM-OCR: Offers general text extraction from images/PDFs, formula extraction to LaTeX, handwriting recognition, and table extraction to Markdown.
• GLM-Image: Provides text-to-image generation capabilities.
• glm-master-skill: A meta-skill for discovering and installing all available GLM skills.

Maintenance & Community

The provided README does not detail specific maintenance schedules, notable contributors, or community channels (e.g., Discord, Slack).

Licensing & Compatibility

This project is licensed under the Apache License 2.0. This permissive license generally allows for commercial use and integration into closed-source projects.

Limitations & Caveats

A ZHIPU_API_KEY is a prerequisite for most skills. Detailed setup and configuration for individual skills may require consulting their respective SKILL.md files, indicating potential complexity beyond the primary installation commands. The README does not specify any alpha status or known bugs.