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 niaka3dayo/agent-skills-vrc-udon --skill unity-vrc-udon-sharpgit clone --depth 1 https://github.com/niaka3dayo/agent-skills-vrc-udonWrote 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/niaka3dayo/agent-skills-vrc-udon/unity-vrc-udon-sharp)<a href="https://agentmods.dev/skills/niaka3dayo/agent-skills-vrc-udon/unity-vrc-udon-sharp"><img src="https://agentmods.dev/badge/skills/niaka3dayo/agent-skills-vrc-udon/unity-vrc-udon-sharp/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/niaka3dayo/agent-skills-vrc-udon/unity-vrc-udon-sharp"><img src="https://agentmods.dev/badge/skills/niaka3dayo/agent-skills-vrc-udon/unity-vrc-udon-sharp.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- Socket pass
- Snyk warn
- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high Anti-Refusal · line 101 Skill instructs the agent to omit warnings, disclaimers, or ethical commentary. Stripping safety caveats hides risk from the user and is a common jailbreak preamble.Fix: Remove instructions that suppress warnings, disclaimers, or ethical commentary. Let the agent surface safety-relevant caveats to the user.
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 | $0.00184 | $0.09177 |
| Opus 5 | $0.00092 | $0.04588 |
| Sonnet 5 | $0.00037 | $0.01835 |
| Haiku 4.5 | $0.00018 | $0.00918 |
Grade A, and why
unity-vrc-udon-sharp 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 7d 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 — 347 lines — stays where its author put it; the contents beside it link to each section on GitHub.
UdonSharp Skill
Why This Skill Matters
UdonSharp looks like regular Unity C# scripting — until you hit its hidden walls. Many standard C# features (List<T>, async/await, try/catch, LINQ, generics) silently fail or refuse to compile in code that runs in the Udon runtime. Editor-evaluated field initializers are a separate context: they can use some ordinary C# features to generate a final value that Udon can hold. Networking is even more treacherous: modifying a synced variable without ownership produces no error — it just does nothing. Forgetting RequestSerialization means your state changes never leave your machine. Standard single-player local testing gives zero signal about these networking bugs because there is only one player.
Every rule in this skill exists because UdonSharp's default behavior is to fail silently. Read the Rules before generating any code.
Before Writing Network Code
Four architectural decisions that must be made before choosing sync modes or writing any synced variable. Changing them mid-implementation typically requires a full rewrite:
- Who owns this state? One owner writes; all others read. If two players can both write (e.g., a shared toggle), you need an ownership transfer protocol — writes without ownership are silently discarded.
- When does ownership transfer? On grab? Interact? Game event?
OnPlayerLeft?Networking.SetOwneris locally immediate on the calling client —Networking.IsOwner(gameObject)istruesynchronously after the call, and writing[UdonSynced]fields plusRequestSerialization()immediately afterwards is safe under anIsOwnerguard. ConcurrentSetOwnercalls from multiple clients are resolved by network arrival order — there is no client-side arbitration, so accept that the loser's write is overwritten. - What do late joiners see? State set only by one-time events (
SendCustomNetworkEvent) is invisible to late joiners. Late-joiner-visible state must live in[UdonSynced]variables, which are delivered automatically viaOnDeserialization; no manualRequestSerialization()on join is needed. - What if the owner leaves mid-session? VRChat automatically transfers ownership to a remaining player (selection rule is not publicly documented), and
OnOwnershipTransferredfires on all clients. Synced variables are preserved, so state is not frozen; decide upfront whether to keep the current value, reset to a known default, or re-apply/re-broadcast derived state inOnOwnershipTransferred.
What ships with it
49 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.
- assets/templates/ArrayUtils.cs 4.6 KB
- assets/templates/BasicInteraction.cs 2.3 KB
- assets/templates/BatchedSync.cs 3.3 KB
- assets/templates/CloggedRetrySync.cs 4.0 KB
- assets/templates/ContactReceiver.cs 8.1 KB
- assets/templates/CustomInspector.cs 9.4 KB
- assets/templates/DataPersistence.cs 12 KB
- assets/templates/DualCopySync.cs 3.1 KB
- assets/templates/EventBus.cs 3.0 KB
- assets/templates/MasterManagedPlayerPool.cs 15 KB
- assets/templates/PackedStateSync.cs 2.8 KB
- assets/templates/PlayerSettings.cs 3.5 KB
- assets/templates/RateLimitedSync.cs 3.6 KB
- assets/templates/StateMachine.cs 12 KB
- assets/templates/SyncedObject.cs 5.6 KB
- assets/templates/UdonSharpProgramAssetAutoGenerator.cs 5.7 KB
- assets/templates/UndoableGameManager.cs 6.0 KB
- CHEATSHEET.md 21 KB
- hooks/validate-udonsharp.ps1 40 KB runs code
- hooks/validate-udonsharp.sh 43 KB runs code
- LICENSE.txt 1.0 KB
- references/api.md 53 KB
- references/assembly-definitions.md 7.3 KB
- references/constraints.md 43 KB
- references/context-preservation.md 10.0 KB
- references/dynamics.md 34 KB
- references/editor-scripting.md 32 KB
- references/events.md 33 KB
- references/image-loading-vram.md 34 KB
- references/networking-antipatterns.md 23 KB
- references/networking-bandwidth.md 14 KB
- references/networking.md 53 KB
- references/patterns-core.md 22 KB
- references/patterns-networking.md 42 KB
- references/patterns-performance.md 45 KB
- references/patterns-ui.md 60 KB
- references/patterns-utilities.md 30 KB
- references/patterns-video.md 28 KB
- references/persistence.md 31 KB
- references/sdk-migration.md 21 KB
- references/sync-examples.md 15 KB
- references/testing.md 10 KB
- references/troubleshooting.md 50 KB
- references/vrctween.md 4.8 KB
- references/web-loading-advanced.md 34 KB
- references/web-loading.md 20 KB
- rules/udonsharp-constraints.md 11 KB
- rules/udonsharp-networking.md 13 KB
- rules/udonsharp-sync-selection.md 3.1 KB
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.
- 7d ago Changed · +3 lines · +12 tokens per session 4c2b92d97b5d
- 12d ago First seen · 344 lines · 172 tokens per session scan A 5ceafd959c24
unity-vrc-udon-sharp is a skill published in the GitHub repository niaka3dayo/agent-skills-vrc-udon (299 stars, last pushed 6d ago), licensed MIT. It adds 184 tokens to every session and 9,177 once invoked, about $0.0009 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 2026-08-30.
Other skills, from other repositories
neo4j-driver-dotnet-skill
Neo4j .NET Driver v6 — IDriver lifecycle, DI registration (singleton), ExecutableQuery fluent API, ExecuteReadAsync/ExecuteWriteAsync managed transactions, IResultCursor (FetchAsync/ ToListAsync), record value access (.Get /As ), null safety, UNWIND batching, temporal types, await using, EagerResult, object mapping…
unity-csharp-architecture-and-ecs
Architectural patterns, Data-Oriented Technology Stack (DOTS), Unity ECS, Burst Compiler, C# Job System, memory management, zero-allocation C# patterns, and frame-budget optimization for high-performance Unity game development.
godot-gdscript-patterns
Enterprise architecture patterns for Godot Engine 4.x using GDScript and C#, Node tree composition, Custom Resources, Signal Bus patterns, Direct Server APIs (PhysicsServer, RenderingServer), memory management, and performance profiling.
avalonia-viewmodels-zafiro
Optimal ViewModel and Wizard creation patterns for Avalonia using Zafiro and ReactiveUI.
csharp-pro
Write modern C# code with advanced features like records, pattern matching, and async/await. Optimizes .NET applications, implements enterprise patterns, and ensures comprehensive testing. Use PROACTIVELY for C# refactoring, performance optimization, or complex .NET solutions.
reflection-method-call
Call a C# method by reflection — including private methods. Requires a method schema obtained via 'reflection-method-find'. Supports static methods, instance methods (with optional target deserialization), and main-thread / off-thread execution.