Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add amyyzing/ai-mcp --skill roblox-mcpgit clone --depth 1 https://github.com/amyyzing/ai-mcpWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/amyyzing/ai-mcp/roblox-mcp)<a href="https://agentmods.dev/skills/amyyzing/ai-mcp/roblox-mcp"><img src="https://agentmods.dev/badge/skills/amyyzing/ai-mcp/roblox-mcp/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/amyyzing/ai-mcp/roblox-mcp"><img src="https://agentmods.dev/badge/skills/amyyzing/ai-mcp/roblox-mcp.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00078 | $0.00909 |
| Opus 5 | $0.00039 | $0.00454 |
| Sonnet 5 | $0.00016 | $0.00182 |
| Haiku 4.5 | $0.00008 | $0.00091 |
Grade A, and why
roblox-mcp scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 11d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 42 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Roblox MCP
Use the connected MCP server's live tool schemas as the source of truth for names, parameters, defaults, and limits.
Core decision rules
Apply these defaults unless the task meets an exception in references/runtime-patterns.md:
- Prefer a specialized read tool over arbitrary Luau.
- Use
search-instancesfor filtered instance discovery. It runsQueryDescendants; do not fetch every descendant and filter it manually. - In custom Luau, use
QueryDescendantsfor selector-expressible instance filters instead of aGetDescendants()scan. - Use
search-gcfor one compact first match. Usegc-snapshot+gc-querywhen the task needs reusable handles, pages, diffs, statistics, or reference traversal. Use rawfiltergconly when neither structured query surface can express the predicate. - Use
get-data-by-codewhen the task needs values returned from Luau. Return compact raw Lua values; do not print data throughexecuteor JSON-encode it manually. - Use
executeorexecute-fileonly for intentional side effects or code that does not need to return data. Verify the effect with a focused read.
Operating workflow
- Select the client. Call
list-clientswhen the target is unclear or multiple clients may exist, then callset-active-clientif needed.- If connection, capability, mapping, or decompiler health is uncertain, call
runtime-status.
- If connection, capability, mapping, or decompiler health is uncertain, call
- Narrow with the cheapest specialized tool:
- One known instance:
inspect-instance; discovery:search-instances, orget-descendants-treewithsummaryOnly=true. - Script metadata:
script-index-status, thenscript-index-startif needed, followed bylist-scripts. - Known text or identifier:
script-grep. - Unknown implementation with known behavior:
semantic-search-scripts. Confirm remote embedding upload only when the configured endpoint is trusted; local Ollama stays local. - Source:
get-script-contentwith a focused line range. - Luraph-protected source: use
operation=runfor an indexed script oroperation=run-sourcefor raw source already in hand. Raw source does not require a Roblox client. Start withcaptureMode=strict, page withoperation=read, release when finished, and retry withsandboxedonly if strict mode stops at an intermediate tree. - Arbitrary compact values:
get-data-by-code. - One GC function/table:
search-gc. Deeper runtime investigation:executor-capabilities, thengc-snapshot,gc-query, andruntime-inspect/runtime-references.
- One known instance:
- Reduce the root, selector, range, filters, and limit before increasing any output budget.
- For GC or custom instance probes, follow references/runtime-patterns.md.
- Mutate only when the user's request authorizes mutation. Use
executeorexecute-file, then verify the resulting state with a specialized read, a targetedget-data-by-codeprobe, or a small console read. - For remotes, call
remote-spywithoperation=list, a small limit, andsummaryOnly=truebefore requesting arguments or changing capture state. - Report the observed result. Top-level execution acknowledgement is not proof that the intended in-game state changed; verify that state separately.
- Use
wait-for-eventcursors for console and non-selector instance journals instead of repeatedly polling snapshots. Selector, attribute, and remote waits observe from call start (or current state withincludeExisting).
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 11d ago First seen · 42 lines · 78 tokens per session scan A 76aacee7aa69
roblox-mcp is a skill published in the GitHub repository amyyzing/ai-mcp (0 stars, last pushed 8d ago), licensed MIT. It adds 78 tokens to every session and 909 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
gameobject-component-destroy
Destroy one or more Components from a target GameObject. Missing (null) components are skipped — they cannot be destroyed. Use 'gameobject-find' and 'gameobject-component-get' to identify the components first.
unity-version-split
Split a C# file into Unity 6.5+ and pre-Unity 6.5 variants. Use when a file needs different implementations for different Unity versions due to API changes (e.g., EntityId vs int, GetEntityId vs GetInstanceID).
godot-signals-groups
Build event-driven, decoupled Godot 4.7 gameplay with signals and node groups: declare and emit custom signals, connect with Callables (incl. bind/one-shot), and broadcast to many nodes via groups and callgroup. Use when wiring node communication in a Godot project, replacing tight references with signals…
motion
How an agent turns a character mesh into a usable animated FBX — and how to judge whether the result is shippable.
unity-addressables
Manage Addressables groups, entries, profiles and content builds (com.unity.addressables, reflection-based).
threejs-exposure-color-grading
Build a measured exposure and grading path in Three.js. Use for a 64x36 encoded luminance meter, asynchronous readback, weighted log-average exposure, asymmetric adaptation, single tone-map ownership, and a generated 32-cube post-tone-map LUT.