Unreal Skill

Control and automate Unreal Editor tasks via OpenClaw AI using HTTP endpoints for level, actor, transform, component, editor, debug, input, asset, console, a...

Security Scan

The skill's code, runtime instructions, and declared requirements are internally consistent with an Unreal Editor integration plugin, but provenance is unclear and there are practical risks (open ports, CORS wildcard, destructive operations) you should consider before installing.

Purpose & Capability

The skill name/description (Unreal Editor control) matches the included code (an HTTP handler that exposes /unreal/* endpoints and a registered 'unreal_execute' tool). There are no unrelated environment variables, binaries, or surprising external services requested.

Instruction Scope

SKILL.md and SKILL_KO.md limit actions to connecting via the OpenClaw Gateway or running an MCP server, registering/polling sessions, queuing commands, reading project logs, and controlling editor state. The docs do include troubleshooting steps like removing /Plugins/OpenClaw/Binaries and /Intermediate (which can be destructive to build artefacts) and recommend running a node MCP bridge from the project tree — both are expected for a plugin but worth caution.

Install Mechanism

This is effectively an instruction-only skill with source code included in the package (extension/index.ts). There is no remote download or extract step in the manifest; nothing in the package attempts to fetch arbitrary code at install time.

Credentials

No environment variables, credentials, or config paths are requested. The code does enable CORS (Access-Control-Allow-Origin: *) and listens on known ports (gateway default 18789 and MCP default 27184), which is functionally necessary but increases network exposure surface if those ports are reachable externally.

Persistence & Privilege

always:false and the skill is user-invocable; disable-model-invocation remains false (agent may call the skill autonomously), which is platform default. Because the skill can perform destructive editor actions (actor.delete, asset.import, editor.play/stop, console.execute), you should consider disabling autonomous invocation or restricting usage if you don't want the model to perform changes without explicit approval.

Assessment

Before installing: 1) Verify provenance — the package has no homepage and the registry owner is an opaque ID; inspect extension/index.ts yourself to confirm behavior. 2) Backup your Unreal project and test in a disposable project first — the docs include rm -rf cleanup commands and the skill can delete or modify actors/assets. 3) Limit network exposure — ensure OpenClaw Gateway (18789) and MCP (27184) are bound to localhost or firewalled so they are not reachable from the public internet; the code sets Access-Control-Allow-Origin: * which can increase risk if ports are exposed. 4) Consider setting disableModelInvocation=true (or otherwise requiring explicit user confirmation) if you want to prevent the agent from autonomously issuing destructive commands. 5) If you decide to proceed, review the full index.ts (already included) for any additional behavior, and run the plugin only in projects where accidental changes are tolerable.

OpenClaw Unreal Plugin

version: 1.0.0

MCP skill for controlling Unreal Engine Editor via OpenClaw.

Connection Modes

Mode A: OpenClaw Gateway (Remote)

The plugin connects to OpenClaw Gateway via HTTP polling. Works automatically when Gateway is running.

Mode B: MCP Direct (Claude Code / Cursor)

The plugin runs an embedded HTTP server on port 27184. Use the included MCP bridge:

# Claude Code
claude mcp add unreal -- node /path/to/Plugins/OpenClaw/MCP~/index.js

# Cursor — add to .cursor/mcp.json
{"mcpServers":{"unreal":{"command":"node","args":["/path/to/Plugins/OpenClaw/MCP~/index.js"]}}}

Both modes run simultaneously.

Editor Panel

Window → OpenClaw Unreal Plugin — opens a dockable tab with:

• Connection status indicator
• MCP server info (address, protocol)
• Connect / Disconnect buttons
• Live log of tool calls and messages

Tools

Level

• level.getCurrent — current level name
• level.list — all levels in project
• level.open — open level by name
• level.save — save current level

Actor

• actor.find — find by name/class
• actor.getAll — list all actors
• actor.create — create actors: StaticMeshActor (Cube, Sphere, Cylinder, Cone), PointLight, Camera
• actor.delete — delete by name
• actor.getData — detailed actor info
• actor.setProperty — set properties via UE reflection system

Transform

• transform.getPosition / transform.setPosition
• transform.getRotation / transform.setRotation
• transform.getScale / transform.setScale

Transform tools require a valid RootComponent (works on StaticMeshActor, PointLight, etc. — not on bare Actor).

Component

• component.get — get component data
• component.add — add component (not yet implemented)
• component.remove — remove component (not yet implemented)

Editor

• editor.play — start PIE (uses RequestPlaySession)
• editor.stop — stop PIE
• editor.pause / editor.resume — pause/resume PIE
• editor.getState — current editor state

Debug

• debug.hierarchy — actor hierarchy tree
• debug.screenshot — capture editor viewport
• debug.log — write to output log

Input

• input.simulateKey — simulate key press
• input.simulateMouse — simulate mouse
• input.simulateAxis — simulate axis

Asset

• asset.list — list assets at path
• asset.import — import asset (not yet implemented)

Console

• console.execute — run console command
• console.getLogs — read project log file; params: count (number of lines), filter (text filter)

Blueprint

• blueprint.list — list blueprints
• blueprint.open — open blueprint (not yet implemented)

Troubleshooting

Stale binaries / plugin not loading

Clear the build cache and restart the editor:

rm -rf YourProject/Plugins/OpenClaw/Binaries YourProject/Plugins/OpenClaw/Intermediate

Connection issues

• Ensure OpenClaw Gateway is running: openclaw gateway status
• Check the Editor Panel log for errors
• Verify the MCP port is not blocked by firewall