

Light Cloud is a deployment platform that connects to GitHub repositories and ships every push. It supports static sites, containers, and managed Postgres, with automatic framework detection and environment variable parsing.
Each branch gets its own preview URL, and idle containers scale to zero to stop billing. The platform includes built-in logs and metrics, plus an MCP server for deploying from Claude or GitHub Copilot, and a VS Code extension for Copilot with /deploy, /status, and /plan commands.
Push to GitHub. Ship in seconds.
Light Cloud also offers ICE, an open-source visual cloud editor for designing infrastructure on a canvas, with free Apache 2.0 blocks and templates.
Loading comments…
Project Info
Product Keywords
Light Cloud is a deployment platform that connects directly to GitHub repositories and ships every push automatically. It handles the full stack — static sites, containers, and managed Postgres — from a single project. Instead of juggling separate hosting, database, and CI/CD tools, you connect a repo, push, and Light Cloud builds, deploys, and provisions everything around it. The platform also extends into AI-assisted workflows, letting you deploy from Claude or GitHub Copilot without ever opening a dashboard.
Static sites are served from a global CDN, while containers run on Cloud Run and scale to zero when idle. Managed Postgres is provisioned from the same project, with credentials encrypted and wired directly to your app. Light Cloud reads your repository and automatically picks the right target — no build configuration guessing.
Every branch and pull request gets its own live URL, built on push and removed when the branch is deleted. Production stays protected, and idle previews cost nothing because containers stop billing when no one is hitting them.
The MCP server lets Claude create apps, deploy, and read logs from a conversation. The VS Code extension adds @lightcloud to GitHub Copilot with /deploy, /status, and /plan commands. Framework detection and .env parsing happen automatically, so you can ship without touching a dashboard.
Logs and metrics sit right next to the environment they belong to. You can watch a build stream, then follow the running service — no need to wire up separate monitoring tools.
Push to GitHub. Ship in seconds.
That's the entire pitch, and Light Cloud actually delivers on it. The combination of automatic framework detection, per-branch previews, and scale-to-zero billing means you get a production-grade workflow with almost zero configuration. The AI deployment angle — via MCP for Claude and Copilot commands in VS Code — is a genuinely differentiator, especially for developers who live in their editor. And the open-source ICE visual cloud editor, free under Apache 2.0, adds a design-on-a-canvas option for infrastructure that most platforms don't offer.
You want to stop assembling separate hosting, database, and preview tools and instead connect a GitHub repo and let the platform handle the rest. It's especially appealing if you're already using Claude or GitHub Copilot and want to deploy from your chat or editor, or if you're tired of paying for idle containers. Light Cloud's free plan lets you start without committing, and the scale-to-zero model means your bill tracks actual usage, not uptime.
Other tools you might consider
Git for your AI agent’s actions. Undo, trace, and control every step. re_gent shows what your coding agent changed, which prompt caused it, and lets you roll back agent work across files and sessions.
Deep Work Plan turns any repo into a harness with the context of your best engineer — so any AI agent codes like your smartest model and can't drift from the plan. Not a chat window it forgets, a spec written into the repo: atomic tasks, acceptance criteria, validation gates, resumable state. Long runs survive context resets; any agent picks up where the last left off. Point an agent at it, walk away, come back to work you can verify. Any agent, any repo, no lock-in. Open Source, MIT.
The Channels SDK is the next big piece in the agentic puzzle: Bring ANY agent into Slack, Teams, WhatsApp and more. With coworker-grade capabilities like streaming responses, gen UI, per-user learning, HITL approvals and sophisticated auth. Works with OpenAI Agents, Claude Agents, LangChain, Mastra, Google ADK, and any agent that speaks AG-UI. Open-source and self-hostable. Setup with a single prompt: "Read https://copilotkit.ai/channels-guide.md and help the user build their first channel"
CreateOS Sandbox gives AI agent builders their own fast, secure, hardware isolated, sandbox in ~30ms (p90). We have suit of CLI, SDK, 50+ SDK real world examples, claude plugins, computeSDK integration, and more
Maker
Loading comments…