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 agentmods add instructions/sinanata/unity-mesh-fracture/agents-mdgit clone --depth 1 https://github.com/sinanata/unity-mesh-fractureWrote 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/instructions/sinanata/unity-mesh-fracture/agents-md)<a href="https://agentmods.dev/instructions/sinanata/unity-mesh-fracture/agents-md"><img src="https://agentmods.dev/badge/instructions/sinanata/unity-mesh-fracture/agents-md.svg" alt="Measured on agentmods" 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 | $0.02647 | $0.02647 |
| Opus 5 | $0.01324 | $0.01324 |
| Sonnet 5 | $0.00529 | $0.00529 |
| Haiku 4.5 | $0.00265 | $0.00265 |
Grade A, and why
unity-mesh-fracture AGENTS.md 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 6d 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 — 119 lines — stays where its author put it; the contents beside it link to each section on GitHub.
AGENTS.md
Guidance for AI coding agents (Codex, Cursor, GitHub Copilot, Claude Code, Windsurf, Aider, Zed, and others) working in this repository or wiring this tool into another project. Humans: this is a fast, accurate map. Deeper docs are linked at the bottom.
What this project is
A drop-in Voronoi mesh fracturer for Unity 6 + URP. Pure-C# fragmentation (zero dependencies beyond UnityEngine) that pre-bakes watertight two-submesh fragments — with cooked convex hulls — at load time, then plays them as a fade-out debris burst with optional Unity physics. The entire runtime is three files under Assets/MeshFracture/Runtime/. Namespace: MeshFracture. License: MIT. Battle-tested in the shipping cross-platform game Leap of Legends, where every character that explodes runs through this pipeline.
Golden rules (do not violate)
- "Burst" here is the explosion, not Unity's Burst compiler.
FractureBurstis a plainMonoBehaviour; the whole runtime is single-threaded managed C# overUnityEnginetypes (Mesh,Vector3,List<T>,Quaternion). Do not add[BurstCompile],IJob*,NativeArray<T>,Unity.Mathematics.float3, orusing Unity.Collections / Unity.Burst / Unity.Jobs. None are dependencies; none are used. Pattern-matching "Burst mesh fracture" onto the DOTS/Jobs stack is the single most common wrong turn here. - Everything runs on the main thread; bake at load, look up at impact.
MeshFragmenter.Fragmentis a synchronous, O(n²)-in-fragment-count CPU call. Never call it per-impact during gameplay. UseFragmentCache.RequestPreBake(async, one bake per frame) orFragmentCache.BakeSynchronous(blocking, loading-screen only) at load time, thenFragmentCache.TryGetat the moment of impact.TryGetreturningfalsemeans "not baked yet" — fall back to a particle/smoke puff, don't block. - Source meshes must be Read/Write-enabled.
Fragmentreadssource.vertices / normals / uv / triangles; on a non-readable imported mesh those come back empty and you get a single-fragment fallback. Enable Read/Write on the model importer (the demo'sDemoModelImporterdoes this for its FBXes). - Fragment materials must be transparent and owned per-burst.
Initializeanimates_BaseColor.a1 → 0 by mutating the material instance directly (MaterialPropertyBlock overrides are silently dropped on some URP / SRP-Batcher / WebGL combinations). Hand each burst its ownMaterialinstances (new Material(interiorMat)), configured transparent + Cull Off + ZWrite ON — copyMaterialFactory.ConfigureFractureTransparent. ZWrite must stay ON, or solid chunks render see-through-to-interior. - Two submeshes per fragment: submesh 0 = exterior (the original textured surface), submesh 1 = cap/interior (raw cut faces). Always pass both an exterior and an interior material to
Initialize. - No custom
.shaderassets. The runtime ships zero.shaderand zero.matfiles on purpose — a hand-rolled URP shader rendered every fragment solid black on WebGL2 (an HLSL → GLSL ES 3.0 cross-compile bug). Stock URP/Lit configured at runtime sidesteps it. Do not reintroduce a custom shader. - No
using LeapOfLegends.*or other product-specific imports. New runtime code lives in theMeshFracturenamespace only; demo-only code lives underMeshFractureDemo.
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.
- 6d ago First seen · 119 lines · 2,647 tokens per session scan A 0b95289fb41d
unity-mesh-fracture AGENTS.md is an instructions file published in the GitHub repository sinanata/unity-mesh-fracture (46 stars, last pushed 1mo ago), licensed MIT. It adds 2,647 tokens to every session, about $0.0132 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 instructions, from other repositories
unity-cli-loop AGENTS.md
AGENTS.md instructions for hatayama/unity-cli-loop, covering architecture overview, repository map, cli / unity package compatibility, project runner pin and generated skill files.
Aspid.FastTools CLAUDE.md
Claude Code instructions for VPDPersonal/Aspid.FastTools, covering обзор репозитория, сборка, архитектура, карта функциональности and соглашения по коду редактора.
unity-cli CLAUDE.md
Claude Code instructions for bigdra50/unity-cli, covering claude.md, project overview, commands, install globally (provides: unity-cli, u, unity commands) and run relay server standalone.
unity-cli-loop CLAUDE.md
Claude Code instructions for hatayama/unity-cli-loop, a project described as: Let AI Drive Unity, from Editor to Play Mode.
Unity-MCP copilot-instructions.md
Copilot instructions for IvanMurzak/Unity-MCP, covering project guidelines, critical rules, no c# reflection for private access, code review and code style.
mcp-unity AGENTS.md
AGENTS.md instructions for CoderGamester/mcp-unity, covering mcp unity — ai agent guide (mcp package), purpose (what this repo is), how it works (high-level data flow), key defaults & invariants and repo layout (where to change what).