5 Patents | 9,500+ API Endpoints | 18 Native DLLs | The AI-Native Spatial Runtime — Now in Early Access Explore the runtime →

Drive RakuAI from your AI assistant.

17 native MCP tools (5 read-only + 12 mutation) plus universal api_call/api_search passthrough to ~9,500 REST endpoints. Production-deployed since March 2026 — now with a versioned public tool contract, a 5-minute read-only connect path, and deny-by-default, audit-logged permissions.

How RakuAI MCP connects an AI assistant to a spatial runtime An AI assistant sends permission-gated, JSON-Schema requests over MCP to the RakuAI runtime, which steps a scene graph and physics simulation and renders to smart glasses, headsets, and the browser. AI assistant Claude · GPT Gemini · Copilot MCP stdio · JSON-Schema permission-gated RakuAI runtime scene graph · physics renders to Smart glasses low-latency display Headset VR / mixed reality Browser WebGL 17 native tools · audit-logged

How it works

Your AI assistant speaks MCP. RakuAI is the runtime on the other end of that connection, driving real spatial output to glasses, headsets, and the browser.

RakuAI MCP architecture: your AI assistant connects over MCP to the RakuAI runtime, which drives smart glasses, headsets and the browser, and exposes 17 native tools plus universal passthrough. Your AI assistant Claude · ChatGPT Gemini · Copilot MCP stdio · JSON-Schema permission-gated RakuAI Runtime scene graph · physics simulation step Output Smart glasses Headset · Browser 17 native MCP tools load_world_model ingest_frame get_scene_state set_render_target start_simulation get_metrics Sandbox / dev All modes 6 of 17 native tools shown

The 17 Native Tools

The full native MCP surface: 5 read-only + 12 mutation tools. Schema-validated, audit-logged, permission-gated. Plus universal api_call/api_search passthrough to ~9,500 raku-api REST endpoints (see spatial-engine.html for the full inventory). The 17-tool surface is now pinned as a versioned public contract (tools.v1.json, schema_version 1) with an anti-drift test that fails CI if the published surface diverges from the live registry.

Tool What it does Permission
get_scene_state(include_physics, include_transforms) Read-only scene-graph snapshot Read-only, all modes
capture_scene(...) Capture a render-time observation (color, depth, segmentation) Read-only, all modes
get_metrics() Performance telemetry (frame time, draw calls, memory, latency) Read-only, all modes
xr_get_session() Inspect headset / glasses session state Read-only, all modes
xr_list_anchors() List spatial anchors in the session Read-only, all modes
add_scene_node(...) Insert an entity into the scene graph Mutation, sandbox/dev
update_scene_node(...) Change properties on an existing node Mutation, sandbox/dev
set_node_transform(...) Position / rotation / scale Mutation, sandbox/dev
remove_scene_node(...) Delete a node Mutation, sandbox/dev
load_world_model(adapter_name, config) Bind a world-model adapter (Runway, Veo, Marble, Genie, content-pack) Mutation, sandbox/dev
ingest_frame(adapter_name, frame_data, frame_index, timestamp) Feed a world-model output frame into the runtime Mutation, sandbox/dev
set_render_target(target_type, config) Configure output (WebGL/VR/native/offscreen) Mutation, sandbox/dev
start_simulation(tick_rate, max_duration, realtime) Step physics + world model under sandbox mode Mutation, sandbox/dev
xr_set_head_pose(...) Drive head pose (sim, recording, or remote) Mutation, sandbox/dev
xr_simulate_input(...) Inject controller / hand-tracking events Mutation, sandbox/dev
xr_create_anchor(...) Anchor a node to world coordinates Mutation, sandbox/dev
xr_remove_anchor(...) Release an anchor Mutation, sandbox/dev

The surface at a glance

17 native tools, grouped by what they're allowed to do. Swipe or use the arrows.

Connect in 5 minutes — read-only

If your model speaks MCP, it speaks RakuAI. Point an existing MCP client at the runtime and get a working read-only session — the 5 read-only tools — on a free sandbox. No model lock-in, no engine SDK, no mutation risk.

Add the server Register raku-mcp in your MCP client config (Claude Desktop, or any MCP-speaking assistant).
Start a read-only session Read-only mode is the default. The 5 read-only tools run in every mode, including production.
Read live scene state Call get_scene_state and get_metrics to see the runtime respond. That delivers immediate value.
Graduate to sandbox The 12 mutation tools unlock in sandbox/dev so you can drive a scene end-to-end.

Step-by-step: the connect-read-only quickstart, the LLM integration guide, and a drive-a-scene walkthrough. On-site: the Claude Desktop guide.

Multi-vendor parity

The contract is identical whether the caller is Claude, ChatGPT, Gemini, or Copilot — that is the wedge. We publish a parity matrix that tracks each vendor against all 17 native tools plus passthrough. Parity holds at the protocol level; client-specific schema handling is tracked openly as a living artifact.

Vendor 17 native tools Passthrough Status
ClaudeProtocol-levelProtocol-levelPublished
ChatGPTProtocol-levelProtocol-levelPublished
GeminiProtocol-levelProtocol-levelPublished
CopilotProtocol-levelProtocol-levelPublished

Honest note: the matrix is published, not yet independently verified. No vendor has been run through a full 17-tool + passthrough conformance sweep on the record — per-client end-to-end verification is open Phase 1 work, tracked in the matrix itself.

The audit & permission story

What lets a security owner approve an autonomous agent acting in space? Mutation is gated by design, every call is logged, and the relay enforces tiered limits.

Deny-by-default All 12 mutation tools are sandbox/dev by default; only the 5 read-only tools run in every mode. Production mutation requires an explicit, audited permission grant.
Per-session audit-log export Every relay call is recorded; the owning account can export its per-session audit trail across the /api/raku/* surface, redaction-safe.
Rate limits & quota Free vs paid relay rate-limit and quota plumbing on the relay surface, so usage is bounded and accountable per tier.

Honest note: the audit export is relay-scoped and account-wide today, not yet per-connector; quota enforcement is active where the relay's counter store is configured. Per-connector granularity and signed audit logs are roadmap.

The MCP spatial tool contract

The 17 tools above are one versioned contract. Same names, same schemas, same permission semantics whether the caller is Claude, ChatGPT, Gemini, or Copilot. The spatial web is standardizing in the open, and this contract is where RakuAI plugs any assistant into it.

We share the full contract with teams evaluating an integration: every tool schema, the permission and audit semantics behind each call, error behavior, and the versioning rules we hold ourselves to. Tell us what you are building and we will send it over.

vs Unity MCP

Unity MCP (just released): scene manipulation, asset placement, build automation — Claude Desktop generates static scenes at design time.

RakuAI MCP (production since March 2026): world-model orchestration — any LLM streams intent into the simulation step every frame.

Unity is the tool your AI picks up. RakuAI is the body your AI inhabits.

Adapters

The contract is the 17 native tools. The adapters are how partners plug in.

VeoEnvironmentAdapter Google's Veo
RunwayEnvironmentAdapter Runway
Custom adapter your model

The adapter contract is public. Today the proven adapters are first-party / reference; a broad third-party adapter ecosystem is a roadmap goal, not the current state.

Why MCP-native matters

  1. No custom integration — if your model speaks MCP, it speaks RakuAI. Connect in read-only mode in about five minutes.
  2. Multi-vendor by design — Claude, ChatGPT, Gemini, and Copilot all target the same 17 native tools; parity holds at the protocol level.
  3. A versioned contract — the 17-tool surface is pinned as tools.v1.json with an anti-drift test, so what you integrate against won't silently change.
  4. Defensible permissions — deny-by-default, audit-logged, schema-validated, with per-session audit-log export for your security review.

What's live vs roadmap

We label what has shipped and what hasn't — honest framing is a feature, not a footnote.

Live today

  • 17 native MCP tools in production since March 2026
  • Versioned public tool contract (tools.v1.json) + anti-drift CI test
  • 5-minute read-only connect path + per-vendor integration guide
  • Published, protocol-level multi-vendor parity matrix
  • Deny-by-default mutation; per-session audit-log export
  • Free vs paid relay rate-limit & quota plumbing
  • Universal api_call/api_search passthrough to ~9,500 endpoints

Roadmap

  • Independently verified per-vendor parity (matrix cells are published, not yet swept)
  • One-click production mutation (today: explicit, audited grants)
  • Per-connector audit granularity and signed audit logs
  • Broad third-party adapter ecosystem (today: first-party / reference)
  • In-product permission-ladder UI; scoped permission profiles
  • Multi-region relay, SSO/SCIM, latency SLOs

Go deeper

Pick a path.