Unreal Engine MCP

There are several Unreal MCP projects on GitHub already, and as of UE 5.8 Epic ships its own experimental first-party MCP plugin (5.8 only, opt-in, requires manually enabling an "Editor Toolset"). Worth being direct about where this project actually differs, rather than just claiming "better":

- Built around reading, not just writing.Most existing projects are strong at creating and manipulating Blueprints from a prompt, but don't address what happens when the model needs tounderstand a large, already-builtproject first. Reading is the first-class citizen here: tiered summaries before full detail, node IDs you can reference without re-fetching.
- A persistent, incrementally-updated project index.The bridge indexes Blueprints, functions, variables, and cross-asset references once, caches it to disk, and updates it fromAssetRegistrydelegates as you edit, instead of rescanning the project on every query.find_referencesanswers "what actually uses this Blueprint" without the model having to enumerate the project itself.
- An optional local-model hook for indexing.If you pointUNREAL_MCP_LOCAL_LLM_URLat a local model (Ollama or anything OpenAI-compatible), indexing summaries are generated there instead of spending Claude's tokens on mechanical scanning work. Fully optional. The index works without it.
- Targets both 5.6 and 5.8 from one codebase, where several existing projects are pinned to a single engine version.

This is being built and verified in public, milestone by milestone. Each milestone's status doc is written honestly, including what's compiled/tested versus what's still unverified:

Both live sessions earned their keep by catching a real bug that no amount of compiling or protocol testing would have surfaced. On 5.8 it wasadd_nodeduplicating an already-present override-event node. On 5.6 it was the.upluginhard-pinningEngineVersionto5.8.0: every build check passed because UnrealBuildTool ignores that field, but the runtime plugin loader honors it, so the editor stopped on a modal incompatibility dialog and the bridge never started.

Since then, all of that has been exercised live too:remove_nodeandVariableGetare covered by the node-id and control-flow suites, andadd_nodenow placesBranch,Sequence,Cast, and standard-library macros (ForEachLoop,WhileLoop, ...) directly, verified by building and compiling a real conditional graph through the bridge alone. Node ids are persistent GUIDs, and every write is undoable with Ctrl+Z under a named "MCP:" transaction. Still outstanding:CustomEvent/VariableSetnode types have not had a dedicated live check, and the M5 catalog coversUFunction-backed nodes; nativeUK2Nodetypes are placed via the dedicatednodeTypevalues rather than discovered throughunreal_find_node.

Ensure you haveNode.js 18+and aUE 5.6 / 5.8project.

Copy theUnrealMCPBridgeplugin folder into your Unreal project'sPlugins/directory:

# macOS / Linux mkdir -p "/path/to/YourProject/Plugins" && cp -r UnrealMCPBridge "/path/to/YourProject/Plugins/" # Windows (PowerShell) New-Item -ItemType Directory -Force -Path "C:\path\to\YourProject\Plugins"; Copy-Item -Recurse UnrealMCPBridge "C:\path\to\YourProject\Plugins\"

Note: Rebuild/open your Unreal project to compile the plugin, and ensure it is enabled in the editor.

There are prebuilt plugin releases on the releases page, and they areolder than this server. The bridge has gained more than twenty commands since the last one, and the protocol number did not change, so an old plugin looks healthy and then fails on the first tool that needs a command it does not have.--doctornow probes for those commands specifically and says so. Building from this checkout is the reliable path.

Install the node dependencies and compile the typescript codebase:

cd mcp-server && npm install && npm run build

3. Check it works before wiring anything up

It reports whether the plugin is reachable, whether its protocol matches the server, whether the project index is built or still scanning, whether the engine's node catalog is readable, and whether a PIE session is in the way. Every failed check comes with the remedy, so you never have to guess which of six things is wrong. Exit code 1 means the editor could not be reached.

Do not hand-write the config.Run this and paste what it prints:

node mcp-server/dist/index.js --print-config # Claude Desktop node mcp-server/dist/index.js --print-config --client cursor # Cursor node mcp-server/dist/index.js --print-config --client claude-code # Claude Code

It emits the exact JSON for this machine, with absolute paths already resolved, and tells you which file it goes in.

That exists because client setup is its own category of failure and all of it is self-inflicted: a missing comma breaks the whole file, a relative path silently does not resolve, and on Windows a barenodemay not be on the PATH the client uses. Every one of those produces the same symptom — the server never starts, with no explanation. The printed config uses the absolute path of the Node that ran the command, so it cannot be the wrong one.

Thenfully quit and reopen the client.Closing the window is not enough, and it is the most common reason a correct config appears not to work.

Point your client at the absolute path ofmcp-server/dist/index.js:

claude mcp add unreal -- node "/path/to/unreal-mcp/mcp-server/dist/index.js"

Claude Desktop(claude_desktop_config.json):

Status: the workflow is committed but has never run.GitHub refused to start it -"the job was not started because your account is locked due to a billing issue"- so no badge is shown here. A badge reading "failing" for a billing reason would say something false about the code. The workflow is structurally valid and the whole suite is verified to pass locally with no editor running, which is the same thing it does on a runner; that is a claim about local runs, not a CI result.

The badge above covers the parts that need no Unreal install, run on a clean Linux machine with nothing preinstalled: build, typecheck, tool/bridge parity, documentation guards, profile token budgets, strict-client protocol conformance, and the unit tests - on the oldest Node the README promises and on the current one. It also checks that--doctorand--print-configbehave correctly withno editor running, since that is exactly when someone reaches for them.

The C++ plugin is not compiled in CI either, because Epic does not ship an engine that can be fetched there.npm run build:enginesbuilds it against every configured engine locally and reports success only if every one of them actually built.

Issues and PRs welcome. This project is young and moving fast, so check the status docs above before assuming something works end-to-end.

This is a web browser that enables your coding agent, such as Claude Code, to visit websites on your behalf and assist you in identifying bugs or creating UI test cases.

next-devtools-mcp is a MCP server that provides Next.js development tools and utilities for AI coding assistants like Claude and Cursor.

A demonstration server for ActionKit, providing access to Slack actions via Claude Desktop.

MCP server that lets Claude Code agents delegate tasks to agents in other project directories, with parallel dispatch, sessions, and async jobs.

Anchor Browser (https://anchorbrowser.io) is secure infrastructure for computer-use agents — stealth cloud browsers, authentication, captcha bypass, and a hosted MCP server for Cursor, Claude, and Windsurf.

Open-source Claude Code plugin replacing Read/Grep/Edit/Bash with token-efficient versions. Independently benchmarked at 57% token reduction on real codebases. 40 MCP tools.

Connects Blender to Claude AI via the Model Context Protocol (MCP), enabling direct interaction and control for prompt-assisted 3D modeling, scene creation, and manipulation.

Context-optimized security plugin for Claude Code that builds architecture maps and redacts secrets.

Charon gives AI clients scoped, auditable access to Lethe memory. connect to Chatgpt or Claude to store memories locally.