A single hub to find Claude Skills, Agents, Commands, Hooks, Plugins, and Marketplace collections to extend Claude Code, Claude Desktop, Agent SDK and OpenClaw

A Python SDK for claude-code hooks

Give Claude Code a memory that evolves with your codebase. Hooks automatically capture sessions, the Claude Agent SDK extracts key decisions and lessons, and an LLM compiler organizes everything into structured, cross-referenced knowledge articles - inspired by Karpathy's LLM Knowledge Base architecture.

A PHP SDK for building Claude Code hooks

A specialized framework for Model Context Protocol (MCP) development featuring 8 Claude Code sub-agents, security hooks, and production-ready FastMCP server templates. Provides immediate MCP development assistance through markdown-driven agents with optional programmatic SDK .

Anthropic Claude Agent SDK for PHP & Laravel — build AI agents with tool use, sandboxing, MCP servers, subagents, hooks, and structured output via the Claude Code CLI

CPM (Claude Project Manager) is an SDK and CLI for managing mono repos with multiple Claude Code projects. It enables sharing skills, agents, hooks, and rules across projects without duplication.

Real-time AI safety guardrails for LLM apps. 10 scanners: prompt injection, PII, harmful content, code vulnerabilities, obfuscation detection. Sub-ms latency. Python + TypeScript SDKs. MCP proxy. Claude Code hooks.

A TypeScript SDK to write claude-code hooks easily with type safety, dependency injection, and comprehensive testing support.

TypeScript SDK for building Claude Code hooks — typed, testable, composable

Canonical BEAM SDK for agentic coding runtimes in Erlang and Elixir. One unified API over Claude Code, Codex, Gemini, OpenCode, and Copilot, with shared sessions, threads, MCP, hooks, telemetry, and backend-native escape hatches.

Curso completo de Claude Code en español: 16 módulos de principiante a experto. Contexto, MCP, hooks, agentes, skills, CI/CD, Agent SDK, plugins y proyecto final. Guía de referencia definitiva.

Claude Code skill for building apps with the Claude Agent SDK — query options, hooks, MCP servers, multi-turn, streaming, permissions, structured output, sessions

Minimal repro: PreToolUse hooks fork the parentUuid chain in multi-turn Agent SDK sessions (anthropics/claude-code#33651)

Mobile-first companion UI for AI coding workflows - Flutter app with OpenCode-like UX, Claude Code Hooks integration, and Agent SDK support

Typed, AOT-compatible C# SDK for Claude Code CLI agent protocol. Port of official Python SDK. Features streaming queries, structured output, stateful conversations, custom MCP tools, permission hooks, and specialized agents. Requires .NET 10+ and Claude CLI.

Build you Claude Code Hooks with Python the easy way

No description available

TypeScript SDK for Claude Code hooks — first typed SDK on npm with testing framework

Demo app showcasing Claude Code hooks, skills, and the Anthropic SDK

Claude Code 환경 설정 도구 모음: hooks, skills, rules, Agent SDK tools

Go SDK for Claude Code CLI — streaming, multi-turn, MCP tool registration, hooks

Claude Code skill for the Claude Agent SDK — build AI agents with query(), hooks, sessions, subagents, MCP, and more

Practice repository for the Hooks and SDK module of the Claude Code in action course.

testing of various claude code techniques- subagents, hooks, commands, github actions, git trees, sdk, mcp etc.

OpenTelemetry instrumentation for the Claude Agent SDK. Auto-generates GenAI semantic convention spans, metrics, and tool traces via SDK hooks — zero code changes required.

Claude Code SDK experiments: custom TUI, in-process MCP tools, hooks, audit logging, and alternative provider support (DeepSeek, GLM, OpenRouter)

demo repo for following the hooks and the sdk section of https://anthropic.skilljar.com/claude-code-in-action/312002

Claude Code skill for building AI agents with the Claude Agent SDK (TypeScript/Python). Covers creating agents, custom tools, hooks, subagents, MCP integration, permissions, sessions, and deployment.

Async Rust SDK for Claude Code CLI — one-shot queries, streaming, multi-turn sessions, permission callbacks, lifecycle hooks, and MCP server support