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 xonovex/platform --skill gpu-rendering-guidegit clone --depth 1 https://github.com/xonovex/platformWrote 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/xonovex/platform/gpu-rendering-guide)<a href="https://agentmods.dev/skills/xonovex/platform/gpu-rendering-guide"><img src="https://agentmods.dev/badge/skills/xonovex/platform/gpu-rendering-guide/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/xonovex/platform/gpu-rendering-guide"><img src="https://agentmods.dev/badge/skills/xonovex/platform/gpu-rendering-guide.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 | $0.00115 | $0.01595 |
| Opus 5.5 | $0.00046 | $0.00638 |
| Sonnet 5.5 | $0.00023 | $0.00319 |
| Haiku 4.5 | $0.00012 | $0.00160 |
Grade A, and why
gpu-rendering-guide 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 1mo 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 — 70 lines — stays where its author put it; the contents beside it link to each section on GitHub.
GPU Rendering Guidelines (Explicit-API Architecture)
API-agnostic architecture for a low-level GPU renderer. The concepts hold across explicit APIs (Vulkan, D3D12, Metal, WebGPU); names differ but the model is the same. For one API's concrete how-to use the matching API skill (e.g. gpu-rendering-vulkan-guide); for the general allocator principle behind GPU memory, see memory-management-guide.
Requirements
- Target an explicit, low-level API where the app owns memory, synchronization, and pipeline state, not a driver that hides them.
- Assume validation/debug layers and a frame debugger are available during development.
Essentials
- Declare, don't sequence - Passes declare resource reads/writes; the graph derives order, barriers, and layout transitions, see references/render-graph.md
- Shaders are build artifacts - Author → compile to a binary intermediate offline → reflect for layouts, see references/shader-system.md
- You own GPU memory - Sub-allocate from a few large blocks; never one allocation per resource, see references/gpu-memory-strategy.md
- You own synchronization - Nothing is implicit; barriers, queue waits, and fences are explicit, see references/synchronization.md
Architecture
- Render graph - Per-frame DAG that prunes dead passes, orders work, and aliases transient targets, see references/render-graph.md
- Shader system - Permutations vs ubershaders, pipeline keyed by shader+state, hot-reload, caching, see references/shader-system.md
- Command recording - One recording context per thread per frame, multi-threaded record, see references/command-recording-and-frames.md
- Sort keys - A 64-bit key per command decouples GPU submission order from CPU record order; merge worker streams and sort before submit, see references/command-recording-and-frames.md
- Programmable vertex fetch - Pull vertices from storage buffers behind a loader interface instead of fixed-function input; skin in-shader from an indirected influence list, see references/vertex-assembly-skinning.md
- GPU-resident simulation - Keep large element state in GPU buffers, advance with compute, and drive draws from a GPU-tracked count via indirect args, no CPU enumeration or readback, see references/gpu-compute-simulation.md
What ships with it
13 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.
- eval-queries.json 8.5 KB
- evals.json 3.4 KB
- moon.yml 166 B
- references/binding-model.md 5.1 KB
- references/command-recording-and-frames.md 4.5 KB
- references/gpu-compute-simulation.md 3.5 KB
- references/gpu-memory-strategy.md 5.5 KB
- references/hdr-output.md 3.8 KB
- references/render-graph.md 4.2 KB
- references/shader-system.md 5.2 KB
- references/synchronization.md 4.8 KB
- references/vertex-assembly-skinning.md 3.7 KB
- SOURCES.md 7.4 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.
- 1mo ago First seen · 70 lines · 115 tokens per session scan A 3c684775ea61
gpu-rendering-guide is a skill published in the GitHub repository xonovex/platform (6 stars, last pushed 4d ago), licensed MIT. It adds 115 tokens to every session and 1,595 once invoked, about $0.0005 per session on Opus 5.5. 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 2026-09-08.
Other skills, from other repositories
harbor-daytona
Use Harbor's Daytona sandbox platform for computer use — creating sandboxes, taking screenshots, sending mouse/keyboard input, and building agent loops. Use when the user wants to interact with a GUI, automate a desktop, do computer use, control a browser visually, or run Claude computer use against a Daytona sandbox.
gamedev-multiplayer
Use when adding multiplayer or netcode to a game — client-server vs P2P, server authority and anti-cheat, state replication vs RPCs, prediction and reconciliation, lag compensation, plus Godot 4 / Unity NGO / Unreal wiring. NOT single-player gameplay (that is godot, unity, unreal), NOT matchmaking or server hosting…
gamedev-physics
Use when working with a game engine's physics — rigid bodies, colliders, collision layers and masks, character controllers, joints, forces versus impulses, raycasts and shapecasts, or when a simulation is unstable (tunnelling, jitter, missed overlaps). Covers the Godot, Unity and Unreal equivalents. NOT rendering or…
hz-unity-meta-movement-sdk-retargeting
Set up and tweak Meta Movement SDK (MSDK) retargeting for a character model. Use this whenever the user wants to retarget a humanoid FBX/prefab for Meta VR body tracking, generate a retargeting config, or hand-edit the resulting .json (fix known-joint mappings, exclude joints from auto-mapping, rename target joints…
swallowtail
Use the complete Godot development loop from the swallowtail CLI: discover the live engine API, build and edit in the open editor, run and play the game, observe state, debug failures, fix them in place, and verify the result. Use when the task involves creating or modifying a Godot project, testing game behavior…
hz-platform-sdk
Guides integration of the Horizon Platform SDK for Meta VR and Horizon OS Android/Kotlin apps — achievements, IAP, users, leaderboards, presence, notifications, abuse reporting, entitlements, asset files, application lifecycle, consent, device integrity, language packs, user age categories, and rate and review. Covers…