Agent-SDK without CLI dependencies, as an alternative to claude-agent-sdk, completely open source

把最丝滑的通用 Agent 体验带进你的工作流，为 100x 专业用户而生的未来产品，正在实现 proactive Agent 阶段。基于 Claude Agent SDK 的完整开源实践，原生支持飞书群聊调用、灵活接入任意大模型供应商 —— 让顶级 Agent 能力真正跑在你每天用的地方。

Blades is a Go-based multimodal AI Agent framework.

Run Claude agents in secure cloud sandboxes — via API, CLI, or Slack. One call. Full agent. Zero infrastructure.

🐔 Agentic systems explained with chickens. Workflows, agents & orchestration made simple. Mermaid diagrams included

Multi-agent orchestration for Claude Code — decompose tasks, coordinate agents, visualize everything in a rich terminal UI

AutoQA-Agent 是一个基于 Claude Agent SDK 构建的开源自动化测试 CLI 工具

Cross-platform Claude Code desktop client by hwwn. Multi-workspace, remote mobile access, session branching, BYO LLM. macOS / Windows / Linux.

Free and open General Purpose Agent

Open source agentic build system - transforms specifications into verified source code with blind adversarial review

Guided labs for the CCA Foundations certification exam. 6 scenarios, 5 domains, 30 tasks.

Control Claude Code from your phone — a multi-machine agent hub via Discord. No API key needed, runs on your Claude Pro/Max subscription. Start new sessions from mobile, manage multiple machines from one place.

Claude Code Plugin instant docs search, interactive HTML courses, and changelog reports. 589 official docs always up-to-date. Auto-discovery Skill, Obsidian & Amber themed courses with animated protocol diagrams, quizzes, and code translations. Zero dependencies.

Lightweight, general-purpose TypeScript agent runtime. Open-source alternative to Claude Agent SDK.

Open-source relational AI framework with identity persistence, memory, and MCP integration. Build relationship-aware AI agents that remember, grow, and maintain continuity. Built on Claude Agent SDK.

Agent runtime with full system prompt control. Define agents in markdown, run them in a terminal TUI or multi-user web UI. Powered by the Claude Agent SDK.

Multi-Agent Continuity Layer — Keep your AI agents on a single, observable, handoff-ready work thread

Study materials, code examples, and practice scenarios for Anthropic's Claude Certified Architect: Foundations (CCA-F) certification exam. Covers all five exam domains: agentic architecture, MCP integration, Claude Code workflows, prompt engineering, and context management.

An Elixir SDK for Claude Code - provides programmatic access to Claude Code CLI with streaming message processing

Swift 6 agent SDK: type-safe tools, streaming, cloud + on-device inference via MLX on Apple Silicon