roblox-studio-mcp
Use when working with Roblox Studio through built-in MCP for scripts, scenes, generated assets, input, or playtesting.
By tabooharmony · 674 installs
npx skills add tabooharmony/roblox-brain --skill roblox-studio-mcp
Source repository · Upstream listing
Roblox Studio MCP
When to Load
Load for Studio MCP work: scripts, scenes, assets, debugging, or playtesting. Skip for standalone code generation.
Quick Reference
Route by capability first. Two Studio MCP bridges exist: the official (built into Studio, closed source) and [chrrxs's](https://github.com/Chrrxs/robloxstudio mcp) (MIT). Route on tools/list ( get connected instances / eval / multiplayer → chrrxs; list roblox studios + per call studio id → official) BEFORE bridge specific bootstrap.
Bootstrap before mutation
1. Identify the target via list roblox studios (official) or get connected instances (chrrxs).
2. Pass the target's studio id (official) or instance id (chrrxs) on every tool call.
3. get studio state and confirm Edit/Client/Server availability.
4. Inspect the target tree and scripts before changing them.
Execution contract
Edit time injection ( multi edit , Edit context execute luau ) writes scripts; it never targets a running playtest. Live evaluation targets a running VM: start play, wait for the intended Client / Server VM, then evaluate there. A stopped VM is never a live evaluation target. Pass datamodel type only where the tool requires it; do not guess from a previous session.
Capabilities
Inspect ( search game tree , inspect instance , script read , script grep ), edit/execute ( multi edit , execute luau ), assets ( search asset / insert asset , generate with wait job finished , store image / upload image ), play/evidence ( start stop play , get console output , screen capture , input simulation, subagent ).
Reliability rules
execute luau is stateless: re acquire references every call.
Read before write; read back after every script, asset, or geometry mutation.
Choose reuse vs generation vs native fallback before asset work.
Call wait job finished before edits that depend on a generation job.
Command size limits are bridge specific. Split large scripts into separate bounded multi edit / execute luau writes (or supported multi edit operations), then read back the full payload to verify.
If MCP is absent, provide offline Luau and state what was not verified.
Full tool mappings, live schemas, asset workflows, and recovery rules: [references/full.md](references/full.md)