My
🤖 Autonomous Cloud Infrastructure

Onboard Your Deployment Agent

Give your AI coding assistants—including Claude, Cursor, AGY, Windsurf, and custom agent loops—the power to autonomously install tools, provision rootless container pods, attach database sidecars, and configure live HTTPS tunnels.

🤖 Claude Code / Anthropic ⚡ Cursor Agent 🚀 Antigravity (AGY) 🧠 ChatGPT / OpenAI 🌊 Windsurf / Roo Code
📋 Copy Agent Prompt ↓ ⚡ View sp CLI Reference →
One-Click Autonomous Agent Setup

Agent Instruction Prompt

📄 View Raw prompt.md

Paste this instruction into your agent chat (Claude, Cursor, AGY, or Windsurf). The agent reads the protocol specification, verifies local tooling, and provisions your application pods non-interactively.

You are my deployment assistant for SiliconPin Cloud. Fetch and follow the autonomous deployment agent instructions from: https://siliconpin.com/prompt.md Your tasks: 1. Verify if the `sp` CLI is installed (`command -v sp`). If not installed, run: Linux/macOS: curl -fsSL https://siliconpin.com/install.sh | bash Windows: irm https://siliconpin.com/install.ps1 | iex 2. Authenticate non-interactively using the environment variable: sp login --token=$SP_TOKEN 3. Inspect this workspace and deploy it as an isolated rootless container pod: sp deploy . --port=3000 4. If a database is required, attach an isolated sidecar: sp deployments attach db mariadb <deploymentId> 5. Report back the public HTTPS deployment link and active status.
Permanent Spec Endpoint: https://siliconpin.com/prompt.md
Compatible with AGY, Claude Code, Cursor Composer, Windsurf

How Agent Onboarding Works

Empower autonomous coding agents in three simple steps.

1 📋

Copy Agent Setup Prompt

Grab the copyable instruction above or pass the instruction file URL directly to your assistant. It specifies all commands, error recovery, and non-interactive flags.

https://siliconpin.com/prompt.md
2 💬

Paste in Agent Chat

Paste into your AI coding tool (Claude Code, Cursor Composer, or AGY session). Export your SiliconPin API token in the environment so the agent can authenticate.

export SP_TOKEN="sp_live_xxx"
3 🚀

Autonomous Execution

The agent checks if sp CLI is present, installs it if needed, creates a rootless container pod, attaches database sidecars, and provisions a public HTTPS domain.

✓ Deployment Active: https://app.example.com
Agent Environments

Quick Integration by AI Tool

How to configure SiliconPin pod deployment in your favorite AI workflow.

🤖

Claude Code (CLI)

Anthropic Terminal Agent

Run Claude directly with the SiliconPin instruction prompt in your repository root:

claude "Read https://siliconpin.com/prompt.md and deploy this repository as a pod with sp CLI"
⚡

Cursor Composer

Agentic IDE Workspace

Open Composer (Cmd+I / Ctrl+I), switch to Agent mode, and paste:

@https://siliconpin.com/prompt.md Follow the instructions to verify sp CLI and deploy this workspace.
🚀

Antigravity (AGY)

Agentic Pair Programmer

Prompt AGY to inspect the current codebase and launch autonomous pod orchestration:

Inspect the codebase, load https://siliconpin.com/prompt.md, and run sp deploy to host this service.
🌊

Windsurf / Roo Code

Autonomous Coding Extensions

Add to your .windsurfrules or project instructions to make SiliconPin deployment default:

# Deployment rules: Follow SiliconPin deployment protocol at https://siliconpin.com/prompt.md using sp CLI.

Agent Command Protocol Matrix

Non-interactive commands executed by autonomous agents in headless environments.

Full sp CLI Docs →
# Install CLI Non-interactively
curl -fsSL https://siliconpin.com/install.sh | bash
# Authenticate via Environment Variable
sp login --token=$SP_TOKEN
# Deploy Source Repository / Folder
sp deploy . --port=3000
# Deploy Container Image
sp deploy api --image=nginx:alpine --port=80
# Attach Database Sidecar
sp deployments attach db mariadb <id>
# Instant Public HTTPS Preview
sp tunnel 3000

Frequently Asked Questions

Everything you need to know about autonomous agent deployment on SiliconPin.

Does the AI agent need sudo or root permissions? +
No. All SiliconPin pods operate within rootless, unprivileged container namespaces. The agent does not require sudo privileges on your workstation or runner.
Where do I get an SP_TOKEN for my agent? +
Log in to your SiliconPin dashboard, navigate to API & Access Tokens, and create a token with deployment permissions. Set it as an environment variable (export SP_TOKEN=your_token) or CI/CD secret.
How does the agent verify that the pod is healthy? +
The agent can run sp my deployments to inspect the active state, health status, and assigned HTTPS URL, or run sp log <deploymentId> to review real-time application logs.
Can I use this in headless GitHub Actions or GitLab CI? +
Yes! The sp CLI supports full non-interactive execution. Check out the ready-to-use GitHub Actions workflow on the sp CLI documentation page.
Can the agent attach persistent databases? +
Yes. The agent can attach MariaDB, MongoDB, or Valkey (Redis) sidecars to any deployment via sp deployments attach db <engine> <deploymentId>, which automatically injects connection credentials into the container environment.

Ready to deploy autonomously?

Copy the agent prompt, set your $SP_TOKEN, and let your AI coding agent manage cloud infrastructure for you.

🚀 Explore Deploy Hub