Product Introduction
- Definition: Sitesolide is a self-hosted, open-source platform-as-a-service (PaaS) and deployment automation tool. It is a command-line interface (CLI) and server stack designed to deploy, manage, and secure web applications on a user-owned Debian 13 server.
- Core Value Proposition: It exists to provide developers and small teams with a Vercel-like deployment experience—simple, isolated, HTTPS-secured deploys—while maintaining full ownership, control, and cost transparency by running on their own infrastructure. Its primary value is self-hosted app deployment, per-app access control, and built-in platform operations without vendor lock-in.
Main Features
- One-Command Deployment: Executing
sitesolide deployfrom a project directory automates the entire deployment pipeline. It builds the application (supporting Bun, Node.js, Python, Go, Ruby, or any binary), restarts it as a systemd service, and performs a health check to verify it responds over HTTPS. This creates an isolated service on the server. - Built-in Authentication & Granular Access Control: Every deployed application sits behind a centralized authentication layer. The platform integrates with any OpenID Connect provider (Google Workspace, Microsoft Entra, Okta) for company sign-in. Administrators can share access to specific apps by individual email, entire domains, or via time-limited guest passwords and access codes. User identity (email and role) is passed to the application with each request.
- Integrated Platform Operations: The server setup includes essential devops tools: Caddy for automatic HTTPS and reverse proxying, systemd for service management and scheduled jobs ("workers"), restic for encrypted hourly backups of project data, and a monitoring system that checks app health every minute and sends alerts (to Slack, Discord, etc.) on downtime/recovery.
- Agent & MCP (Model Context Protocol) Support: The CLI is designed for automation, offering a
--jsonflag on all commands for machine parsing. It includessitesolide mcp, a server that integrates directly with AI coding assistants like Claude Code, Cursor, and Codex, allowing these agents to perform deployments and management tasks programmatically.
Problems Solved
- Pain Point: The complexity and cost of deploying and securing multiple small applications, prototypes, or internal tools. Traditional methods involve manual server configuration, SSL setup, and access management, while hosted PaaS solutions can be expensive and lock data into a third-party platform.
- Target Audience: Indie developers and small technical teams who build many apps (e.g., with AI assistants) and need a simple, private way to host them. Agency developers who need to deploy client projects on isolated subdomains with client-specific access. Internal tooling teams that require secure, access-controlled staging and sharing of internal applications.
- Use Cases: Deploying a Claude Code or Cursor-built prototype to a live, shareable URL in seconds. Hosting a private internal CRM or dashboard accessible only to company employees via their corporate SSO. Providing a client preview of a web project using a time-limited password. Automating scheduled background jobs (workers) as part of an application stack.
Unique Advantages
Strengths & Limitations (Pros & Cons):
- Pros: Full infrastructure ownership and data control (self-hosted). Transparent cost (only server + domain). No per-seat licensing; the MIT license allows unlimited users and projects. Built-in, robust authentication and granular sharing. Comprehensive built-in ops (backups, monitoring, HTTPS). Excellent for AI-assisted development workflows via MCP.
- Cons: No zero-downtime deployments; the app restarts during
deploy. No high-availability or failover (single-server architecture). Requires user management of server kernel updates and reboots. Backups are initially local until configured for an external bucket. Less abstraction than container-based solutions, tied to Debian/systemd.
Key Alternatives & Differentiation:
- Vercel: A fully-managed, cloud-hosted platform. Sitesolide differentiates by being self-hosted (data on your server), having no per-seat cost, and offering built-in, app-specific access controls without an Enterprise plan. Vercel offers global CDN and serverless functions, while Sitesolide offers simpler, full-stack app hosting on one machine.
- Coolify: A self-hosted PaaS that heavily utilizes Docker. Sitesolide differentiates with its simpler architecture (no Docker, uses systemd/Caddy), its deep integration of company sign-in and per-app sharing as a core feature, and its native MCP support for AI agents.
- Kamal (formerly MRSK): A deployment tool from Basecamp for running containerized apps on multiple servers. Sitesolide differentiates by being a full platform (including auth, backups, monitoring) rather than just a deployment tool, and by focusing on single-server, non-containerized simplicity for a portfolio of smaller apps.
Frequently Asked Questions (FAQ)
- How much does sitesolide cost? Sitesolide itself is free and open-source under the MIT license. You only pay for the underlying cloud server (VPS) and domain name, giving you full control over hosting costs and no platform fees.
- Can I use sitesolide with a custom domain or client domains? Yes. Sitesolide automatically manages DNS to provide each app with a subdomain (e.g.,
app.yourdomain.com). It also supports using completely separate custom domains (e.g., a client's domain) and will create the necessary DNS records automatically with supported providers like Cloudflare, Route 53, and others. - How does authentication and security work? All apps are placed behind a central Caddy proxy that handles HTTPS termination and authentication. You configure an OpenID Connect provider (like Google Workspace). Users log in once at the platform level, and their access is governed by per-app permissions set by the admin. The apps themselves receive the user's verified email and role.
- Is sitesolide suitable for production applications? It is designed for production-like robustness (HTTPS, isolation, backups, monitoring) for small to medium internal tools, prototypes, and client projects. However, its architecture lacks zero-downtime deploys and built-in failover, which may be critical for large-scale, public-facing commercial applications.
- What is the "MCP server" and how do AI agents use it? The Model Context Protocol (MCP) is a standard for AI assistants to interact with tools. The
sitesolide mcpserver exposes deployment and management functions. This allows AI coding assistants like Claude Code or Cursor to directly run commands likedeployorshareon your behalf, streamlining the build-and-ship workflow.
