Most software engineers are seriously sleeping on how good LLM agents are right now.

Once you've got Gemini Code set up, you can point it at your codebase and have it operate like a super-powered teammate. The real unlock is building a solid set of Skills and Agents for the stuff you do all the time.

For background on how this project was produced, see the companion notes repo: https://github.com/joeangel/gemini-showcase-notes

This project demonstrates specific "Magic" capabilities powered by Gemini:

See an issue? Just comment:

@gemini /fix

Gemini will analyze the issue, write the code to fix it, and open a PR for you.

Open a PR, and Gemini automatically reviews it against our Code Review Agent checklist (Security, Types, UI States).

Open an issue, and Gemini immediately analyzes it and applies the correct labels.

Want to modernize your code? Comment:

@gemini /refactor convert this to async/await

Gemini acts as a Senior Engineer, using our specialized Smart Diff tool to safely restructure your code.

Have a big idea but don't know where to start?

@gemini /plan implement user login

Gemini acts as a Project Manager, breaking down your request into a detailed task.md checklist.

Mention @gemini in any comment to ask questions or request changes.

Detailed documentation has been moved to keep this README token-friendly for AI agents.

Before you start, please read our Cost & Usage Guide.

• GitHub Account: Free Tier is fine, but be aware of Action Limits.
• Gemini API Key: Get one from Google AI Studio.

• Add Secret: Go to Settings -> Secrets -> Actions and add GEMINI_API_KEY.
• Permissions: Enable Read and write permissions in Settings -> Actions -> General.

mkdir -p .gemini/{agents,commands,hooks,skills}

Create GEMINI.md in your project root. See Example GEMINI.md.

Create .gemini/settings.json. See Example settings.json.

MIT

Notice: Portions of this software are derived from ChrisWiles/claude-code-showcase, copyright (c) 2025 Chris Wiles, licensed under the MIT License.