Connect Agents · MCP + agent skills

Connect your AI client.

Give your agent the tools to manage your games, and the instructions to package and publish them. Choose your client for its exact configuration.

MCP server

The connection that reads and changes your game entries.

MCP server URL

https://aigamelabs.net/mcp

Streamable HTTP · OAuth sign-in and consent · no shared API key

Publishing skill

The workflow for metadata, artwork, WebGL packaging, Unity exports and Unreal limitations.

Publishing skill URL

https://aigamelabs.net/mcp/skills/SKILL.md

One SKILL.md file · installing it does not authorize an agent

29 client guides

Choose your client

The configuration and skill instructions below change with this selection.
For clients with a documented project skill directory.

Client configuration

AnythingLLM

OAuth bridge requiredSkill used as reference

Desktop app or self-hosted Docker. Remote MCP servers are configured in anythingllm_mcp_servers.json, but only static headers are documented — no built-in OAuth — so aigamelabs needs the third-party mcp-remote stdio bridge on Desktop. No SKILL.md import: its "custom agent skills" are Node.js plugins.

Step 1 · Tools

Connect the MCP server

  1. AnythingLLM documents no OAuth flow for remote MCP servers — only type sse/streamable entries with url + static headers — so https://aigamelabs.net/mcp cannot authenticate natively; a headerless streamable entry will not pass OAuth.
  2. On AnythingLLM Desktop (v1.8.0+), use the mcp-remote stdio bridge instead. Node.js 18+ must be installed on your machine first: AnythingLLM does not install commands for you and requires the command on PATH.
  3. Edit plugins/anythingllm_mcp_servers.json inside your AnythingLLM storage directory (it is created when you open the Agent Skills page) and merge in the aigamelabs entry below. On first start the bridge runs the OAuth flow (DCR + PKCE) in your browser and caches tokens in ~/.mcp-auth.
  4. Start the server from the Agent Skills page (gear icon → Start) or invoke @agent. AnythingLLM loads MCP tools only — no Resources, Prompts, or Sampling.
  5. Docker deployments: do not use this bridge there — see the warning below. Per-account OAuth is not available in AnythingLLM either way.

Choose the example for your platform and merge the aigamelabs entry into the existing configuration. Keep unrelated servers and settings.

anythingllm_mcp_servers.json — Desktop only: mcp-remote bridge (OAuth handled by the bridge)

plugins/anythingllm_mcp_servers.json (inside your AnythingLLM storage directory)
// Merge the "aigamelabs" entry into the existing "mcpServers" object.
// Desktop only — an mcp-remote bridge in a shared Docker deployment would
// hold one shared login for every user of that instance (see notes).
{
  "mcpServers": {
    "aigamelabs": {
      "command": "npx",
      "args": ["-y", "mcp-remote", "https://aigamelabs.net/mcp", "--transport", "http-only"]
    }
  }
}

Before you connect
  • Do not run the bridge in AnythingLLM Docker: the mcp-remote OAuth token cache (~/.mcp-auth) would live once inside the container and be shared by everyone using that instance — that is a server-wide shared login, not safe multi-user per-account auth. Keep the bridge on Desktop, where the token cache belongs to one local user.
  • mcp-remote is a third-party proxy package (npm: mcp-remote, MIT, Node 18+), not an AnythingLLM component; the OAuth behavior in this profile comes from the package README, not from AnythingLLM documentation. Remove the bridge if AnythingLLM ships native OAuth for remote servers.
  • Only the sse/streamable entry format (url + static headers) is documented by AnythingLLM itself; there is no documented native connection for this OAuth-protected server.

Step 2 · Workflow

Add the publishing skill

  1. AnythingLLM's extension point is Node.js "custom agent skills" (plugin.json + handler.js, written in JavaScript) — there is no documented SKILL.md/Agent Skills import or auto-loading.
  2. Download https://aigamelabs.net/mcp/skills/SKILL.md and use it as reference reading when composing prompts, or as the specification if you choose to hand-write a custom agent skill for the publishing workflow.
  3. The aigamelabs tools (add_game, update_game, remove_game, list_games, get_game, upload/release, …) remain callable via @agent once the bridge is connected; the skill file stays documentation — it is never installed as a skill.

Download for reference (not auto-loaded by AnythingLLM)

curl -fL -o SKILL.md https://aigamelabs.net/mcp/skills/SKILL.md

Download SKILL.mdRead the skill

Reading or attaching the Markdown is a workflow reference, not automatic native skill discovery. It does not connect MCP or grant account access.

After setup

Try a read-only first request

First prompt for your agent

Read the aigamelabs-game-publishing skill, then use AIGameLabs list_games to show my game entries. Do not create, publish or remove anything yet.

Once the connection works, the agent can use add_game, update_game and remove_game, plus build, artwork and release tools. Removing a game archives it and revokes delivery; publishing still requires an approved release. Publication is manual by default; explicitly opt in when freezing a release only if you want that exact snapshot to publish automatically on approval.

Your approval remains in control

Sign in on aigamelabs.net and inspect the client's name, callback and requested scopes. Grant games:read for reading and games:write for changes; offline_access permits refresh tokens. Never paste browser cookies, Google credentials or an access token into shared configuration. Creator access, upload limits and human review still apply.

Manage or disconnect an agent in Settings → Connected agents. A recent sign-in is required to disconnect. Review game metadata and ask for confirmation before destructive actions, rights attestations or automatic-publication opt-ins.

Connection help

The URL shows “missing authorization header”

That is the expected OAuth challenge, not a page to sign in on. Add the URL in your client's MCP settings, then use its Connect or Authenticate action to open the consent flow.

The sign-in link expired or the callback was rejected

Reconnect from the MCP client to request a fresh authorization link. Do not bookmark a signed consent URL. Older clients that omit the native application type are inferred from loopback or private-use callbacks; explicitly configured web clients must use HTTPS callbacks. Unsafe callbacks remain rejected.

The client has HTTP support but cannot authenticate

HTTP transport alone is insufficient: this server requires OAuth. Follow the selected guide's native OAuth or documented bridge instructions. A stdio bridge runs locally and stores its own authorization state; do not share that cache or use one person's login for a multi-user server. This server uses dynamic client registration, not URL-based client metadata discovery.

The skill is not listed

Keep the exact filename SKILL.md inside the directory shown for your client, preserve its frontmatter, then reload or start a new session as that client's instructions require. A reference-only app will not show it as an installed native skill.

A game operation is denied after connection

OAuth grants access only to your own workspaces. Creator enrollment, quotas and human review gates remain in force. Read-only consent cannot authorize writes. Ask for the required scope rather than copying a token into configuration or trying to bypass review.