🤖 AI Agents Weekly: Gemini CLI, Qodo Gen CLI, Context Engineering, Claude Apps, AlphaGenome

Gemini CLI, Qodo Gen CLI, Context Engineering, Claude Apps, AlphaGenome

In today’s issue:

• Google releases Gemini CLI

• Qodo Gen CLI allows you to build custom dev agents

• OpenAI adds both deep research and webhooks to their APIs

• Context engineering is the new prompt engineering

• Anthropic now allows you to build and deploy interactive apps

• Google DeepMind introduces AlphaGenome

• ElevenLabs launches 11ai, a voice-first AI personal assistant

• Cursor now allows you to review PRs and leave comments

• New course on building LLMs from Scratch

• Top AI dev news, research, and much more.

---

---

Top Stories

Gemini CLI

Google has launched Gemini CLI, a free and open-source AI tool that brings the power of Gemini 2.5 Pro directly to the command line. Designed for developers, it delivers fast, local, and flexible access to AI assistance without leaving the terminal. Gemini CLI supports a broad range of use cases, from code generation and debugging to content creation and task automation, and it integrates seamlessly with Google’s Gemini Code Assist.

Key highlights:

• Free, Generous Usage: Developers using personal Google accounts can access Gemini 2.5 Pro with a 1M token context window, 60 model requests per minute, and 1,000 per day, entirely free during preview.

• Powerful Capabilities: Built-in tools allow users to ground prompts via Google Search, extend functionality through the Model Context Protocol (MCP), and automate tasks non-interactively within scripts.

• Shared Tech with Code Assist: Gemini CLI shares core technology with Gemini Code Assist, enabling multi-step reasoning agents in both terminal and IDE (VS Code) environments.

• Fully Open Source: Released under Apache 2.0, Gemini CLI encourages community contributions, with customizable configurations and support for system prompts and shared team settings.

Blog | GitHub