AI agents
Run your server with an AI agent.
AlphaGSM ships 35 skills for Cursor, Claude, Codex, and other agents that can read a SKILL.md file and use a terminal. You name the server and the result. The agent uses the same commands you would type.
Point Cursor, Claude, or Codex at the repository
Open the AlphaGSM repository in the agent and give it a terminal on the host where AlphaGSM is installed. Say which game, which server name, and what result you want. If the install is a standalone binary or a config file outside the checkout, give that path too.
For server management, the first words should name alphagsm-agent-operations. For a code change, name alphagsm-agent-development. Those entry skills choose the narrower skill. You do not need to load all 35.
- Codex discovers skills through
.agents/skills/, which links to the files underskills/. - Cursor, Claude Code, and other SKILL.md agents can be told to read SKILLS.md and the selected skill file when automatic discovery does not list them.
- On a Windows checkout, Git may turn those links into text files. Tell the agent to read
SKILLS.mdand the skill path directly. Native discovery needs working directory links.
The upstream walkthrough is Run AlphaGSM With An AI Agent. This page is the catalogue those agents and search crawlers can read without opening the repository.
Entry points
Name one entry skill. It selects the narrower skills for the request.
| Skill | Use it when | Instructions |
|---|---|---|
alphagsm-agent-operations | Run or manage AlphaGSM servers from a broad operational request. | Skill file |
alphagsm-agent-development | Implement, fix, or maintain AlphaGSM through a verified repository change. | Skill file |
Server operations
These skills operate a server that already has a name, a host, and a result you can check.
| Skill | Use it when | Instructions |
|---|---|---|
alphagsm-server-provision | Create, set up, or start a named AlphaGSM game server without interactive prompts. | Skill file |
alphagsm-host-preflight | Check whether a host and installation can run a selected AlphaGSM server. | Skill file |
alphagsm-runtime-operations | Select or diagnose process versus Docker execution for an existing AlphaGSM server. | Skill file |
alphagsm-server-settings | Change AlphaGSM game settings, secrets, map selection, or port values. | Skill file |
alphagsm-content-operations | Apply supported server mods, plugins, addons, or maps through AlphaGSM desired state. | Skill file |
alphagsm-world-operations | Create a missing game world or reset explicitly requested configured world data. | Skill file |
alphagsm-console-admin | Read AlphaGSM server logs, send game-console input, or broadcast player messages. | Skill file |
alphagsm-health-check | Verify actual game readiness, query or info health, or complete server shutdown. | Skill file |
alphagsm-incident-response | Respond when AlphaGSM setup or start crashes, hangs, times out, or regresses on a selected server. | Skill file |
alphagsm-backup-restore | Make or restore an AlphaGSM backup with an explicit server and recovery target. | Skill file |
alphagsm-server-upgrade | Upgrade game server files or the AlphaGSM manager, then verify recovery and health. | Skill file |
alphagsm-server-migration | Move an AlphaGSM server between hosts with verified data preservation and a deliberate cutover. | Skill file |
alphagsm-boot-activation | Enable or disable AlphaGSM server startup on boot using the existing user scheduler. | Skill file |
alphagsm-fleet-operations | Operate an explicitly selected group of AlphaGSM servers and report each server's outcome. | Skill file |
Repository development
These skills change the AlphaGSM repository. They do not operate a live server unless the request also names one.
| Skill | Use it when | Instructions |
|---|---|---|
alphagsm-module-authoring | Add a new dedicated-server game module with lifecycle and runtime coverage. | Skill file |
alphagsm-server-enablement | Fix a disabled or skipped game server and close an AlphaGSM support-tracker backlog item. | Skill file |
alphagsm-config-sync | Wire supported set values into real game configuration or fix lifecycle config drift. | Skill file |
alphagsm-port-claims | Implement or debug AlphaGSM port ownership, collision checks, and claim-bearing edits. | Skill file |
alphagsm-provider-requirements | Implement provider credentials, tokens, licenses, provisioning, or an AUTH or BYO override path. | Skill file |
alphagsm-curated-content | Add curated multiplayer server content or fix shared payload staging and ownership. | Skill file |
alphagsm-test-triage | Diagnose AlphaGSM lint, unit, integration, smoke, or CI failures from actual diagnostics. | Skill file |
alphagsm-runtime-validation | Verify AlphaGSM runtime families, branch-local images, and process or Docker lifecycle parity. | Skill file |
alphagsm-release-preparation | Prepare validated AlphaGSM versions, standalone binaries, checksums, or release workflow changes. | Skill file |
alphagsm-skill-maintenance | Author or validate AlphaGSM skills, discovery links, routing, and behavioral evaluations. | Skill file |
Repository contracts
These skills keep module, runtime, documentation, and release changes inside the repository's existing rules.
| Skill | Use it when | Instructions |
|---|---|---|
changelog-discipline | Update AlphaGSM release notes when release-relevant behavior or workflow expectations change. | Skill file |
disabled-server-gate | Diagnose the disabled creation gate or change a module support state. | Skill file |
docker-runtime-wiring | Implement or review explicit Docker runtime hooks in an AlphaGSM game module. | Skill file |
install-layout | Change game install paths, archive extraction, executable lookup, or config file layout. | Skill file |
server-info-gathering | Implement or debug AlphaGSM query and info protocol hooks. | Skill file |
server-lifecycle | Change game module lifecycle hooks or check the AlphaGSM command contract. | Skill file |
smoke-driven-docs | Update server guides or command examples from verified AlphaGSM smoke workflows. | Skill file |
system-install-validation | Investigate a downloaded game server that cannot start or answer queries on its installed host. | Skill file |
wiki-publishing | Publish the wiki or change the repository documentation mirror when that is the explicit request. | Skill file |
Example requests
- Use alphagsm-agent-operations. Create Minecraft Vanilla as mymc in /srv/alphagsm/mymc on port 25565. Start it, check that it answers, and leave it running.
- Check mytf2. Explain why its game query fails and repair the cause.
- Back up mymc and tell me which archive was created and what it includes.
- Install MetaMod and SourceMod on mytf2, then verify that the server loads them and answers queries.
- Update first and second, keep them running afterward, and report each result.
- Enable mymc on reboot, but leave it stopped today.
- Use alphagsm-agent-development. Fix this integration failure and open a pull request on the latest default branch.
Repeat a request against an existing server name. The agent should inspect and reuse that server rather than create another one or overwrite saves. Name the archive when restoring. Name the server and world when resetting progress. Mod installation and the Workshop commands are the content steps those requests should follow.
What the agent checks and reports
The operations skills check the selected host, configuration, and game requirements before changing anything. They use AlphaGSM for the lifecycle. A running process is not treated as a healthy server. The report should include the running state, the readiness check, any backup or recovery artifact, and any remaining blocker. Fleet commands are reported per server.
A repository change follows the lifecycle, runtime, and support contracts, updates tests, docs, and the changelog when the change requires them, and includes the validation that was actually run. Preparing a release is a different request from publishing it.
What these skills will not do
- They will not supply a Steam password, a game server login token, a license, or owned game files. AUTH modules need your login. BYO modules need the files, URL, or local service the guide names. The agent should stop and tell you which prerequisite is missing.
- They will not put credentials into the prompt, the logs, or a shared report. Use the installation's private configuration path.
- They do not start a resident monitoring agent. Ongoing checks need a scheduler you explicitly request.
- They do not replace the host tools. The agent still needs permission to run AlphaGSM, Docker, or the game's runtime.
Everyday commands, without an agent, are in Server operations.
Frequently asked questions
Which skill should I name?
Name alphagsm-agent-operations to run a server. Name alphagsm-agent-development to change the AlphaGSM source. The entry skill picks the focused skill, such as backup, health check, or incident response.
Does the agent need the AlphaGSM website?
No. It needs the repository or the installed CLI, plus a terminal on the host. This page publishes the catalogue so agents and search indexes can find the skill names and their limits.
Will it type my Steam password?
No. Skills explain the missing AUTH or BYO requirement. They do not invent credentials or ask you to paste them into the prompt.
Can it watch the server forever?
Only if you explicitly ask for a scheduler. The skills themselves do not launch a persistent agent service.