Ai Build Tools

Forge

From product spec to live URL — scaffold a backend, ship it to your VPS.

Forge is a CLI and agent-skill ecosystem that turns plain-English product requirements into deployed, production-grade backends. Describe your business in a sentence, and Forge's pipeline handles everything from schema design to GitHub Actions auto-deploy on your VPS.

Instead of manually wiring every skill, MCP server, and agent rule yourself, the forge CLI installs and configures the ecosystem your AI coding tools need. The same setup that scaffolds your data platform also provisions CI/CD, GHCR image builds, and Cloudflare DNS — so "describe and deploy" is one continuous flow.

It powers outputs across the full delivery chain:

• schema design & migrations
• repository and service code
• integration tests
• frontend data contracts
• GitHub Actions CI/CD workflows
• GHCR container builds & VPS auto-deploy
• Cloudflare DNS provisioning
• runbook helpers (bin/logs, bin/rollback)

What Forge actually is

Forge is an ecosystem bootstrapper.

It helps you:

• detect supported AI agents/editors
• install the Forge orchestrator skill and its sibling skills
• configure MCP servers like Datafy
• verify that everything is wired correctly

Once installed, your configured AI agent can use that ecosystem to generate real backend artifacts — and deploy them — from a business or product specification.

From spec to live URL in 10 minutes

Forge works in two phases that together form an end-to-end pipeline:

1. Scaffold with forge — Describe your business. Forge (via the b2dp skill) generates schema, services, migrations, tests, and frontend data contracts. Dockerfile and CI workflow are included.

1. Ship with ship-to-vps — The ship-to-vps skill turns your GitHub repo into a live deployment: pushes the Docker image to GHCR, runs the CI pipeline on push, configures Caddy + Cloudflare DNS, and provides runbook commands for logs and rollbacks.

The forge init command configures both phases in a single pass. You describe your product once; Forge and ship-to-vps handle the rest.

The 8 skills

| Skill | Description | |---|---| | forge | Converts business specs into a fully provisioned data platform and product implementation | | ship-to-vps | Ships any web app to a Hetzner-class VPS via GHCR, GitHub Actions, Caddy, and Cloudflare DNS | | api-test-generator | Generates comprehensive API integration tests from existing backend repos and schemas | | cloud-solution-architect | Cloud architecture design following Azure Architecture Center best practices | | context7-mcp | Library, framework, and API reference lookup for setup questions and code generation | | frontend-data-consumer | Scaffolds typed React/Vue data components from backend API contracts | | frontend-design-review | Reviews and creates production-grade frontend interfaces with design system compliance | | infrastructure-as-code-architect | Translates local infra config into Terraform/Pulumi IaC for AWS, Azure, or GCP |

Supported environments

The CLI is designed to work with supported agent/editor setups such as:

• Claude Code / Claude Desktop
• Cursor
• VS Code / Copilot-style workflows
• Gemini CLI
• Antigravity

Quick start

Install

npm install -g @teckedd-code2save/forge

Run setup

npx @teckedd-code2save/forge init

If you installed it globally, the same setup is available as:

forge setup

Or accept detected defaults automatically:

forge init --yes

To scope configuration to the current repository:

forge init --project

Verify configuration

forge check

Explore installed skills

forge skills list
forge skills info forge

1. Command: npx @teckedd-code2save/datafy@latest --config /path/to/your/dbhub.toml --transport stdio
2. Click Save and reload the window (or restart Cursor).

forge skills

Inspect available skills and view skill details.

Datafy MCP setup

The ecosystem relies heavily on @teckedd-code2save/datafy, which gives AI assistants database-aware capabilities such as schema inspection, SQL execution, and code generation.

Example MCP config:

{
  "mcpServers": {
    "datafy": {
      "command": "npx",
      "args": [
        "@teckedd-code2save/datafy@latest",
        "--config",
        "/path/to/your/dbhub.toml",
        "--transport",
        "stdio"
      ]
    }
  }
}

Adding to Codex

1. CLI: Use codex mcp to add and manage servers.
2. Config: Update ~/.codex/config.toml (global).
3. TOML Configuration:

[mcp_servers.datafy]
command = "npx"
args = ["@teckedd-code2save/datafy@latest", "--config", "/path/to/your/dbhub.toml", "--transport", "stdio"]

Including the forge Skill

> Forge turns product requirements into backend delivery workflows.

If you want to use this skill, ensure the forge/SKILL.md file is placed in the specific directory monitored by your AI assistants:

• Claude: Place skills in ~/.claude/skills/
• Gemini CLI: Place skills in ~/.gemini/skills/
• Antigravity: Place skills in ~/.gemini/antigravity/skills/ or .agent/workflows/ in your workspace.
• Codex: Place skills in ~/.agents/skills/ (global) or .agents/skills/ (project).

Repository

GitHub: https://github.com/teckedd-code2save/ai-build-tools

License

MIT

Source & license

This open-source MCP server is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: teckedd-code2save
• Source: teckedd-code2save/ai-build-tools
• License: MIT
• Homepage: https://teckedd-code2save.github.io/ai-build-tools/

Install and usage instructions live in the source repository linked above.