Agent Skills Hub

goodqunbot

AI-powered WeChat message assistant. Retrieve recent messages from contacts/groups and send messages directly via AI chat. Windows 10/11 only.

exa

Neural web search and code context retrieval via Exa AI. Ideal for documentation, technical research, code examples, and company intelligence.

refactoring

Linter-driven refactoring agent that resolves complexity issues like cyclomatic depth, primitive obsession, and long functions using automated pattern extraction.

moonbit-refactoring

Refactor MoonBit code for idiomatic style: minimize public APIs, implement methods, leverage pattern matching with views, and enforce high test coverage for maintainable MoonBit projects.

explaining-code

Explains code using visual diagrams, relatable analogies, step-by-step walkthroughs, and common pitfalls.

documenting-rust-code

Standardized Rust documentation practices for the HASH codebase, ensuring consistency in doc comments, intra-doc links, and error handling.

skill-orchestrator

Orchestrates complex programming tasks by analyzing available skills, generating structured execution plans, and managing manual or delegated multi-step workflows.

structured-outputs-advisor

Expert advisor for implementing Anthropic's structured outputs. Choose between JSON mode and strict tool use for guaranteed schema compliance and validated agentic workflows.

agentic-workflows

Build production-grade AI agents using LangGraph, Anthropic/OpenAI/vLLM, and structured outputs. Features streaming, A2A protocol, Pydantic validation, vector memory, and guardrails for resilient, multi-agent workflows.

deepwiki-rs

AI-powered documentation engine that automatically generates C4 architecture diagrams, technical specs, and codebase analysis from any source code directory.

fabric

Intelligent pattern selection for Fabric CLI, automatically choosing from 242+ specialized prompts for threat modeling, data analysis, summarization, and content creation.

hegelion

Dialectical reasoning and adversarial coding agent for MCP-enabled editors, forcing LLMs to resolve internal contradictions for higher quality outputs.