Better and Open-Source Claude Code

Today’s top AI Highlights:

1. Better than Claude Code, and it's open-source

2. Vercel’s open-source full-stack vibe coding platform

3. Claude can now create Excel sheets, Word docs, PDFs, PPTs

4. Claude Code and OpenAI Codex are coming to Bolt

5. Launch, switch, and manage Claude Code, Codex, Gemini CLI from one terminal UI

& so much more!

Read time: 3 mins

AI Tutorial

We have created a complete Google Agent Development Kit crash course with 9 comprehensive tutorials!

This tutorial series takes you from zero to hero in building AI agents with Google's Agent Development Kit.

What's covered:

• Starter Agent - Your first ADK agent with basic workflow

• Model Agnostic - OpenAI and Anthropic integration patterns

• Structured Output - Type-safe responses with Pydantic schemas

• Tool Integration - Built-in tools, custom functions, LangChain, CrewAI, MCP

• Memory Systems - Session management with in-memory and SQLite storage

• Callbacks & Monitoring - Agent lifecycle, LLM interactions, tool execution tracking

• Plugins - Cross-cutting concerns and global callback management

• Multi-Agent Patterns - Sequential, loop, and parallel agent orchestration

Each tutorial includes explanations, working code examples, and step-by-step instructions.

Everything is 100% open-source.

We share hands-on tutorials like this every week, designed to help you stay ahead in the world of AI. If you're serious about leveling up your AI skills and staying ahead of the curve, subscribe now and be the first to access our latest tutorials.

Don’t forget to share this newsletter on your social channels and tag Unwind AI (X, LinkedIn, Threads) to support us!

Latest Developments

Better Than Claude Code, and It's Open-Source 💪🔓

This alternative to Claude Code is not just open-source, it's better! Codebuff is a multi-agent CLI coding assistant that beats Claude Code by 61% vs 53% on over 175 real-world coding tasks.

Codebuff deploys a team of specialized AI agents that work together through an Orchestrator. A file explorer agent scans your codebase, a planner agent creates the implementation strategy, and implementation agents make the actual changes while review agents validate everything.

The breakthrough came when the team behind this agent refactored their "sprawling spaghetti code" into just half a dozen TypeScript files, each defining a single agent.

Now they have open-sourced both the CLI coding tool and the underlying AI agent SDK. Use the Codebuff CLI agent if you’re building products using AI agents, or tinker with their open-source SDK to build your own custom AI agents.

Key Highlights:

1. For the AI Coder - Codebuff offers a ready-to-use AI coding agent that runs in your terminal. It's designed for daily development tasks and serves as a practical, open-source alternative to tools like Claude Code.

2. For the AI Builder - The underlying TypeScript SDK allows you to create your own multi-agent applications with programmatic control and model flexibility.

3. Agents as composable units - Each agent defines specific capabilities and spawnable subagents, creating modular workflows where file pickers find relevant code, planners create implementation strategies, and reviewers validate changes.

4. Model agnostic - Works with any model on OpenRouter, including Anthropic, OpenAI, Google, and open models like Qwen or DeepSeek, allowing optimal model selection for different agent roles.

Vercel’s Open-Source Full-Stack Vibe Coding Platform 🪄🖥️

Vercel just released an open-source end-to-end vibe coding platform, heavily inspired by v0, that can generate full-stack applications from simple text prompts. It can write and read files, run commands, install packages, autofix errors, and everything else.

The platform creates secure Linux sandboxes that support any programming language and runtime, from Node.js and Python to Go and Rust.

Every action streams in real-time - you can watch the AI think, see files being generated, monitor command execution, and get live preview URLs of running applications, all while errors are automatically detected and fixed in the background.

Key Highlights:

1. Complete Linux Dev Environment - Full-featured sandbox that supports any programming language and runtime, with file explorer, command terminal, live preview, chat interface, and real-time error monitoring built with Next.js and AI SDK.

2. Advanced Tool Orchestration - Sophisticated tool calling system with parallel execution, dependency management, and automatic state recovery across sandbox sessions.

3. Multi-Model - Choose between GPT-5, Claude 4 Sonnet, Grok, Amazon Nova Pro, Google Gemini, and other models through Vercel AI Gateway. GPT-5 includes configurable reasoning effort levels for higher effort for more complex coding tasks.

4. Streaming Architecture - Real-time UI updates for every aspect of development: file generation progress, command execution status, error detection, and live application previews.

Your boss will think you’re a genius

If you’re optimizing for growth, you need ecomm tactics that actually work. Not mushy strategies.

Go-to-Millions is the ecommerce growth newsletter from Ari Murray, packed with tactical insights, smart creative, and marketing that drives revenue.

Every issue is built for operators: clear, punchy, and grounded in what’s working, from product strategy to paid media to conversion lifts.

Subscribe for free and get your next growth unlock delivered weekly.

Sign up now

Quick Bites

Claude can now create and edit files aka Claude Code Interpreter 
Claude can now create actual files - Excel spreadsheets, documents, PowerPoint decks, and PDFs - directly within the chat interface. Users can upload raw data and receive polished outputs with working formulas, charts, and analysis, or describe what they need built from scratch. The feature runs on a private computer environment where Claude writes and executes code behind the scenes. Currently available for Max, Team, and Enterprise users.

Claude Code and OpenAI Codex are coming to Bolt
Bolt just announced they're bringing Claude Code, OpenAI Codex, and Gemini into their visual web builder, becoming the first platform to orchestrate existing coding agents instead of rolling their own. The strategy is smart: let the frontier labs handle what they do best while Bolt focuses on seamless integration and user experience.

Tools of the Trade

1. DevTeam CLI - Run a team of local coding agents in your terminal. Launch multiple Claude Code, Codex or Gemini agents, switch between them, review their changes and add comments, and push PRs all from one streamlined terminal UI.

2. Whisper - A desktop AI assistant that continuously monitors your screen activity, documents, and interactions to build contextual memory and assistance you proactively, whether you’re researching, attending a meeting, or coding. It can also take actions across apps via MCP.

3. Dev3000 - This tool by Vercel automatically monitors your web app development session by capturing all server output, browser interactions, network activity, and visual states into timestamped logs. The tool integrates with AI assistants through MCP servers for debugging with full context.

4. Awesome LLM Apps: A curated collection of LLM apps with RAG, AI Agents, multi-agent teams, MCP, voice agents, and more. The apps use models from OpenAI, Anthropic, Google, and open-source models like DeepSeek, Qwen, and Llama that you can run locally on your computer.
   (Now accepting GitHub sponsorships)

Hot Takes

1. Local models are a waste of time

   Change my mind ~
   Matt Shumer
   
   

2. all vibe coded websites look identical ~
   hayden

That’s all for today! See you tomorrow with more such AI-filled content.

Don’t forget to share this newsletter on your social channels and tag Unwind AI to support us!

Unwind AI - X | LinkedIn | Threads

Awesome LLM Apps | Sponsor Us

PS: We curate this AI newsletter every day for FREE, your support is what keeps us going. If you find value in what you read, share it with at least one, two (or 20) of your friends 😉