Zed X Gemini Agent: This NEW AI Agent in Zed AI Editor is ACTUALLY AMAZING!

Key Concepts

• Zed: A fast, open-source code editor built in Rust.
• Gemini CLI: Google's command-line interface for their Gemini AI models.
• Agent Client Protocol (ACP): Zed's protocol for integrating AI agents into the editor.
• MCPs (Micro Code Plugins): Plugins that extend the functionality of AI agents.
• Ninja Chat: An all-in-one AI platform offering access to multiple AI models.

Zed's Integration of Gemini CLI

• Deep Integration: Zed now offers first-class support for Gemini CLI, integrating Gemini's models directly into Zed's Rust-based environment. This integration provides the speed and responsiveness of Zed with the features of Gemini CLI.
• Graphical Interface: Gemini CLI is presented in a simple sidebar within Zed, eliminating the need to switch between the terminal and the editor.
• In-Place Code Generation and Refactoring: Users can generate and refactor code directly within the editor and receive instant answers from the AI agent.
• Real-Time Agent Following: The agent's actions can be followed in real-time, with changes presented in a review interface.
• Agent Client Protocol (ACP): Zed has created an ACP, allowing any agent to integrate into Zed and provide a graphical interface. This enables community or company-developed agents to be compatible with Zed.

Using Gemini CLI in Zed

• Updating Zed and Gemini CLI: Users need to update Zed to the latest version and upgrade Gemini CLI with ACP support.
• Creating a New Thread: In the agent panel, users can create a new thread with Gemini CLI.
• Referencing Files, Rules, and Web Pages: The integration allows referencing files, rules, and web pages within the agent.
• Follow Agent Option: Enabling the "follow agent" option allows users to see the agent's work with diffs.
• Example Demo: The video demonstrates using Gemini CLI to create a simple Minesweeper game, showcasing the speed and responsiveness of the integration.

MCPs and Rules

• MCP Configuration: MCPs configured in Zed's agent do not automatically transfer to Gemini CLI. Users need to configure MCPs in the Gemini CLI config file.
• MCP Marketplace: Zed has an MCP marketplace, but MCPs need to be manually installed into Gemini CLI.
• Context 7 and Bite Rover: The video mentions using Context 7 (for documentation) and Bite Rover (for memory layer) MCPs. Bite Rover allows users to make and retrieve memories that are synced across coder agents and shared with teammates.
• Rules: Similar to MCPs, rules are not synced between Zed's agent and Gemini CLI and need to be manually created.

Limitations and Missing Features

• Slash Commands: Slash commands are not interactive, meaning suggestions don't appear when typing the first letter.
• Undocumented Features: Some features mentioned in Zed's blog post, such as highlighting code for Gemini and generating functions from comments with hotkeys, were not found to be working or configurable.

Ninja Chat Advertisement

• All-in-One AI Platform: Ninja Chat is presented as an all-in-one AI platform with access to models like GPT40, Claude 4, Sonnet, and Gemini 2.5 Pro for $11 per month.
• Features: The platform includes an AI playground for comparing model responses and a mind map generator.
• Pricing: The basic plan includes 1,000 messages, 30 images, and five videos monthly.
• Discount Codes: The video provides discount codes "king25" for 25% off any plan and "king40yearly" for 40% off annual subscriptions.

Conclusion

The integration of Gemini CLI into Zed is a significant step towards a more seamless AI-assisted coding experience. While the integration is still in its early stages and has some limitations, it offers a promising glimpse into the future of code editors. The Agent Client Protocol (ACP) opens up possibilities for community-driven agent development, allowing users to choose the best agent for their needs. The video highlights the potential of this integration while also acknowledging the areas that need further development and refinement.