Now open for building: Introducing Gemini CLI extensions (blog.google)

🤖 AI Summary
Google has launched Gemini CLI extensions — an open framework that lets developers customize the Gemini CLI agent by installing pre-packaged integrations that bring external tools directly into the terminal. The move plugs popular services (Dynatrace, Elastic, Figma, Postman, Shopify, Snyk, Stripe and more) into an ecosystem that already counts over one million Gemini CLI users. Extensions are discoverable on a new marketplace page (ranked by GitHub stars) and can be installed from a GitHub URL or local path with a single command: gemini extensions install <GitHub URL or local path>. Technically, each extension bundles a “playbook” that teaches the model how to use specific tools, plus optional MCP (Model Context Protocol) servers, context files (GEMINI.md or custom types), excluded tools and custom slash commands. At runtime Gemini CLI consults the playbook and local context (files, git status, environment) to choose and run the right tool, enabling agentic workflows like real-time debugging, CI/CD remediation, API automation and security scanning without context switching. Google also ships extensions for cloud-native workflows (Cloud Run, GKE, gcloud), app building (Code Review, Flutter, Firebase) and generative/data tooling, making the CLI a composable, intelligent hub for developer productivity.
Loading comments...
loading comments...