# InsForge

> InsForge is an open-source backend platform built for AI coding agents, exposing databases, auth, storage, and functions through a semantic layer agents can understand and operate.

InsForge is an open-source backend development platform designed specifically for AI coding agents and AI code editors. It acts as a semantic layer between agents and backend primitives — including databases, authentication, storage, edge functions, and a model gateway — so agents can understand, reason about, and operate full-stack applications end to end. InsForge is available as a cloud-hosted service at insforge.dev or as a self-hosted instance via Docker Compose, with one-click deployment options for Railway and Zeabur (Sealos coming soon).

**Key Features:**

- **Semantic Layer for Agents** — *InsForge performs backend context engineering, exposing structured schemas and documentation so AI agents can fetch context, configure primitives, and inspect backend state without manual intervention.*
- **Authentication** — *Built-in user management, authentication flows, and session handling, configurable directly by agents via the MCP server.*
- **PostgreSQL Database** — *A fully managed Postgres relational database with pgvector support for embeddings and vector search.*
- **S3-Compatible Storage** — *File storage compatible with the S3 API, accessible and configurable by agents.*
- **Model Gateway** — *An OpenAI-compatible API gateway that routes requests across multiple LLM providers.*
- **Edge Functions** — *Serverless code execution running on the edge via Deno, deployable by agents.*
- **Site Deployment** — *Integrated site build and deployment pipeline for full-stack app shipping.*
- **MCP Server Integration** — *Connect InsForge to any MCP-compatible AI coding agent (e.g., Cursor, Claude Code) to give it full backend access through a single prompt.*
- **Open Source (Apache 2.0)** — *The full platform source code is publicly available on GitHub, free to use, modify, and distribute.*

**Self-Hosted Setup (Docker Compose):**

Prerequisites: Docker and Node.js installed on your machine.

1. Clone the repository and enter the directory:

   ```
   git clone https://github.com/insforge/insforge.git
   cd insforge
   ```

2. Copy the example environment file:

   ```
   cp .env.example .env
   ```

3. Start the services with Docker Compose:

   ```
   docker compose -f docker-compose.prod.yml up
   ```

4. Open `http://localhost:7130` in your browser and follow the dashboard prompts to connect the InsForge MCP Server to your AI coding agent.

5. Verify the setup by prompting your agent: *"I'm using InsForge as my backend platform, call InsForge MCP's fetch-docs tool to learn about InsForge instructions."*

**One-Click Deployment:**

Deploy a pre-configured InsForge instance on Railway or Zeabur without installing Docker locally (Sealos support coming soon).

**Cloud-Hosted Plans:**

For managed hosting, InsForge offers a Free tier for prototypes and side projects, a Pro tier ($25/month) for production apps that scale with included compute and AI model credits, and a custom Enterprise tier with SOC2, SSO, and HIPAA available as a paid add-on.

## Features
- Semantic layer for AI coding agents
- PostgreSQL relational database with pgvector
- S3-compatible file storage
- Built-in authentication and user management
- OpenAI-compatible model gateway across multiple LLM providers
- Edge functions via Deno
- Site build and deployment
- MCP server integration
- Self-hosted via Docker Compose
- Multi-project isolation
- One-click deployment on Railway, Zeabur, Sealos
- Real-time and WebSocket support
- Open source under Apache 2.0

## Integrations
Docker, PostgreSQL, Deno, OpenAI API, pgvector, Railway, Zeabur, Sealos, Cursor, Next.js, OAuth2

## Platforms
WEB, API, CLI

## Pricing
Freemium — Free tier available with paid upgrades

## Version
v2.0.8

## Links
- Website: https://insforge.dev
- Documentation: https://docs.insforge.dev/introduction
- Repository: https://github.com/InsForge/InsForge
- EveryDev.ai: https://www.everydev.ai/tools/insforge