Menu
 

AI Game Server Hosting With Supercraft: Tutorial for Agents and Builders

How to Use Supercraft With AI Agents

Supercraft is not a generic AI tutorial topic. It is a practical platform for using AI agents to sell, provision, and manage game servers. An agent can browse plans, recommend the right fit, generate a secure checkout or login link, claim authenticated access after the user approves, and then continue into real operational work such as installs, config updates, server restarts, logs, and console commands.

The key point: this is about AI use of Supercraft, not “AI” in the abstract. That makes it useful both for real builders and for searchers who want a concrete workflow instead of vague automation claims.

What an AI Agent Can Do With Supercraft

Workflow Stage What the Agent Does Main Endpoints
Discovery Lists games, compares plans, explains features, and helps the user choose the right option. GET /catalog/games
GET /catalog/games/{game_code}
GET /catalog/regions
Approval Creates a checkout flow for new users or a magic login link for existing customers. POST /orders/purchase
POST /auth/login-link
Authentication Claims a JWT after the user completes payment or approves the login flow. GET /auth/claim?magic=...
Operations Installs the game, edits config, restarts the server, reads logs, and sends console commands. /servers and sub-endpoints

Step 1: Start With Catalog Discovery

The best AI workflow starts with plan discovery, not configuration. Supercraft exposes a dedicated catalog layer so an agent can understand the offer before recommending anything to the user.

GET https://claws.supercraft.host/catalog/games
GET https://claws.supercraft.host/catalog/games/vlh
GET https://claws.supercraft.host/catalog/regions

This lets the assistant reason about:

  • Which game the user wants
  • How many players they expect
  • Whether they need mod support
  • Which region is closest to their group
  • Whether a short or longer billing period makes sense

Step 2: Recommend a Plan With Real Data

When the agent reads a game’s catalog entry, it gets the plan matrix, pricing periods, features, and product IDs required for checkout. That means it can move from a vague recommendation like “get a mid-sized server” to a specific one like “Plan M in Europe because you want 8 players and mod support.”

Before the sale

Use plan and region data to help the user buy the right server, not just any server.

After the sale

Use the same integration to continue into install, config, restart, logs, and support work.

Step 3: Generate a Checkout or Login Link

Supercraft supports two clean user-approval flows:

  • New user checkout: POST /orders/purchase
  • Existing user login: POST /auth/login-link
POST https://claws.supercraft.host/orders/purchase
Content-Type: application/json

{
  "product_id": 1234
}

The purchase response includes a checkout_url and a short-lived magic token. The login flow works similarly for existing users. This is one of the strongest parts of the product design: the user still approves access explicitly, but the agent can continue the workflow after that approval without forcing the customer into a manual dashboard-only experience.

Important implementation rule: the assistant should send the checkout or login URL to the user before it starts polling the claim endpoint. This is explicitly called out in the official getting-started guide because tool calls can block message delivery in many agent runtimes.

Step 4: Claim the JWT

After the user completes payment or clicks the magic login link, the integration polls:

GET https://claws.supercraft.host/auth/claim?magic=TOKEN_HERE

When the response changes to status: ready, the system returns a Bearer JWT. From there, the agent can act on the user’s behalf against the authenticated server endpoints.

Step 5: Install the Game and Pick the Right Region

Once authenticated, the agent can list the user’s servers, identify the new pending instance, and install the requested game into the best region.

GET /servers
POST /servers/{id}/install

{
  "game": "vlh",
  "region": "ew",
  "branch": "stable"
}

This makes Supercraft useful beyond the pre-sales story. The assistant can continue from recommendation to actual deployment, which is exactly what many “AI hosting” products fail to support.

Step 6: Use the API for Real Operations

Task Endpoint Why It Matters
List servers GET /servers Find the right server and understand current deployment state.
Inspect status GET /servers/{id} See whether the server is pending, installing, started, or stopped.
Get connection details GET /servers/{id}/connection Return host, port, password, and join info back to the user.
Read current config GET /servers/{id}/config Inspect live settings before making changes.
Read config schema GET /servers/{id}/config/schema Lets the agent make schema-aware changes instead of brittle text edits.
Update settings PUT /servers/{id}/config Apply structured configuration changes safely.
Restart or stop POST /servers/{id}/restart, /stop, /start Complete everyday admin tasks after config edits or troubleshooting.
Logs and console GET /servers/{id}/logs, POST /servers/{id}/console Support debugging, moderation, and scripted in-game actions.

Step 7: Prefer Schema-Aware Config Updates

One of the best parts of the Supercraft agent API is the config schema endpoint. A good assistant should not guess config keys. It should:

  1. Read /servers/{id}/config/schema
  2. Read /servers/{id}/config
  3. Prepare a minimal update payload
  4. Write changes with PUT /servers/{id}/config
  5. Restart if the game requires a reload

That is a better AI workflow than editing raw config files blindly. It reduces errors, makes validation possible, and gives support or ops agents a safer surface to work with.

Best Uses for AI on Supercraft

  • Sales assistants that compare plans and generate checkout links
  • Support copilots that inspect logs, status, and connection details
  • Internal ops tools that standardize installs, config changes, and restart flows
  • Hosted agent integrations that use the OpenAPI spec and public docs as the implementation surface

Best Practices

  • Always send the user-facing checkout or login URL before polling for the claim.
  • Persist the JWT securely after a successful claim.
  • Use X-Agent-Key for trusted integrations.
  • Read the config schema before proposing changes.
  • Confirm destructive operations such as restarts or game switches.
  • Use logs and connection endpoints instead of asking the user to manually fetch technical details from the panel.

Documentation to Pair With This Guide

If you are evaluating or implementing AI workflows around Supercraft, start with these resources:

Why This Topic Works for SEO

Searchers looking for “AI game server hosting,” “agent API for game servers,” or “how to manage game servers with AI” want something practical. Supercraft has a real answer: plan discovery, user-approved auth, purchase flows, installs, structured config, logs, console actions, and post-sale operations. That combination makes this a strong tutorial topic with both technical depth and commercial relevance.

Next step: start with the /ai page, then move into the agent guide and API docs when you are ready to build or test an integration.

Top