create-toolset
Skill16 starsChanged 3 months agoMITView on GitHub
Use when authoring or extending an Unreal Engine Toolset registered with ToolsetRegistry and exposed through Unreal MCP. Trigger for new AI-callable C++ or Python tool methods, ToolsetRegistry registration, schemas, converters, and Toolset tests; skip invoking existing tools or authoring an Unreal Agent Skill.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- EpicGames/unreal-engine-skills-for-claude-code-plugintheir own version, first seen Jun 7, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- EpicGames/unreal-engine-skills-for-claude-code-plugin★ 1.8ksame name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository2 other files it gives its agents.
- unreal-mcpSkill
- unreal-skillSkill
The file
We split this skill into the rules it gives an agent
9 rules. About 0 are common, 6 are distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈6
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 10
Use unreal-mcp instead when the task only invokes tools that already exist.
- Very specific: line 19
If the Editor is connected, use Tool Search to inspect existing Toolsets.
- Distinctive: line 24
Add to an existing Toolset when its domain already fits.
- Distinctive: line 38
Composable: use consistent real UE types so results from one tool can feed another without string parsing.
- Distinctive: line 42
All exposed methods must be static.
- Distinctive: line 49
Document the Toolset domain, every tool, parameters, return meaning, units, ranges, encodings, and empty/null semantics that cannot be …
- Distinctive: line 54
Every tool requires a success-path test and coverage for each explicit error condition.
- Very specific: line 59
Prefer Python when all required APIs appear in <project>/Intermediate/PythonStub/unreal.py.
- Distinctive: line 71
Before implementing tests or reporting completion, read references/testing.md.
≈ “Common” is an estimate: at least 1 in 5 of the 364,980 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Create Toolset
Author or extend an Unreal Engine Toolset: a collection of static, AI-callable
functions registered with ToolsetRegistry and exposed through Unreal MCP.
Use unreal-mcp instead when the task only invokes tools that already exist.
Establish scope
Before editing code:
- Confirm the exact project, plugin, module, and allowed files. Do not modify an installed Engine checkout or enable project settings without explicit user authorization.
- If the Editor is connected, use Tool Search to inspect existing Toolsets. If
it is not connected, search nearby
Toolsetsfolders and read relevant headers or Python modules. Do not add functionality that is already exposed. - Decide whether the capability is general. Prefer an existing generic
Toolset—such as
AssetToolsorObjectTools—over a domain-specific duplicate. - Add to an existing Toolset when its domain already fits. Create a new Toolset or plugin only when ownership is distinct; ask the user when that placement decision is not clear.
- Compare Python and C++ API coverage, explain the tradeoff, and let the user choose when either is viable.
Design contract
A good Toolset is:
- Clean: expose the smallest useful API rather than mirroring a complex UE subsystem directly.
- Complete: provide sensible create/read/update/delete symmetry. Read-only getters are fine when mutation is not meaningful.
- Composable: use consistent real UE types so results from one tool can feed another without string parsing.
- Non-duplicative: reuse generic Toolsets and shared helpers.
All exposed methods must be static. Use real parameter and return types rather than JSON encoded inside strings. A normal return communicates success; an exception or UE script error communicates failure. Do not use boolean returns, error strings, or result wrappers merely as status channels. Async UE result objects are the exception because they explicitly represent eventual success or failure.
Document the Toolset domain, every tool, parameters, return meaning, units,
ranges, encodings, and empty/null semantics that cannot be inferred from the
signature. Remove descriptions that only repeat names, types, defaults, or
obvious call order. Use UPROPERTY metadata for machine-readable constraints.
Every tool requires a success-path test and coverage for each explicit error condition. Do not declare the Toolset complete from schema discovery alone.
Choose an implementation
Prefer Python when all required APIs appear in
<project>/Intermediate/PythonStub/unreal.py; search the stubs narrowly rather
than loading the entire file. If stubs are missing, tell the user that generating
them requires enabling Python Developer Mode and restarting the Editor—do not
change that setting silently. Choose C++ when Python coverage is materially
insufficient. Stop and report the gap if neither surface exposes the required
API rather than inventing a fragile workaround.
- For C++ declarations, registration, async results, converters, and errors, read references/cpp.md.
- For Python decorators, registration, reload, and errors, read references/python.md.
- Before implementing tests or reporting completion, read references/testing.md.
Handoff
Report the Toolset and plugin changed, language chosen, AI-callable methods,
registration path, tests run, actual MCP discovery result when available, and
any restart/reload requirement. Keep UE runtime mutations within the user's
explicit authorization and follow the unreal-mcp Skill when using live Editor
tools for verification.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

