Browse the library

Context Mode is an MCP server that optimizes AI coding agent context windows by sandboxing tool output and persisting session memory, significantly reducing token usage.

A lightweight, secure AI assistant that runs agents in isolated containers with long-term memory, connecting to various messaging apps.

An MCP server that acts as an IDE for coding agents, providing semantic code retrieval, editing, refactoring, and debugging tools to improve agent efficiency and reliability in complex codebases.

agentmemory

rohitg00

A plugin offering persistent memory for AI coding agents like Claude Code and Cursor, extending their knowledge base with confidence scoring, knowledge graphs, and hybrid search.

A vast open-source library of 380+ production-ready skills and plugins compatible with Claude Code, Cursor, and other AI coding agents for various domains.

A plugin that provides persistent file-based planning for AI coding agents, ensuring task plans and progress survive context loss and enabling robust long-running tasks.

A marketplace of 100+ agentic skills, commands, and plugins for product management, covering discovery, strategy, execution, launch, and growth for AI agents like Claude Code.

A standard specification and documentation for creating portable agent skills, enabling AI agents to gain specialized knowledge and repeatable workflows.

A curated collection of over 100 specialized Claude Code subagents designed for various development tasks and easily installable as plugins.

The official Python SDK for Model Context Protocol (MCP), enabling developers to build custom MCP servers and clients to extend AI agent capabilities.

agency-agents-zh

jnMetaCode

A collection of 277 plug-and-play AI expert roles for various departments, supporting multiple AI coding tools and enabling team collaboration.

A collection of practical AI skills for agents like Claude Code and Codex, providing utilities for goal definition, disk analysis, information retrieval, and code hygiene.

SkillOpt optimizes and trains natural-language skills for LLM agents using a machine learning-like approach, producing deployable best_skill.md artifacts.

ARIS provides Markdown-only skills for autonomous ML research, enabling cross-model review loops and experiment automation for various AI agents.

A local proxy to use any LLM like Gemini, DeepSeek, or Ollama with agents like Claude Code and Cursor.

An open-source platform for building and deploying durable AI agents and workflows in TypeScript, featuring long-running tasks, retries, and queues.

claude-obsidian

AgriciDaniel

Enables Claude Code to build a self-organizing Obsidian knowledge base from any source, creating linked Markdown notes and answers grounded in your vault.

An open-source desktop application for building and managing a multi-agent AI workforce with MCP and skill integration, featuring browser and terminal toolkits.

Local-first search engine and MCP server that provides AI agents with code search, memory, and RAG capabilities.

A curated list of Claude Skills, resources, and tools for customizing Claude AI workflows, especially for Claude Code.

A tool that converts various sources like documentation sites, GitHub repos, and PDFs into structured knowledge assets and AI skills for platforms like Claude.

Provides a lightweight, agent-driven personal memory system that acts as a shared LLM wiki across different agents and devices.

ccstatusline

sirmalloc

Add a customizable status line to Claude Code CLI displaying model info, Git status, and token usage.

Track AI coding token usage and costs across tools like Claude Code and Cursor by model, project, and task.