vgpu is a TypeScript library that provides one WebGPU programming interface for browser rendering, headless Node execution, and deterministic tests, including shader modules, 3D scenes, GPU tensors, neural networks, and mathematical visualizations. Developers use it to write explicit GPU drawing and computation code while sharing it across runtimes without hidden global state. The catalogue skill supports work with this library.
Borrowing it
Nothing to install: this file belongs to vercel-labs/vgpu. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/vercel-labs/vgpu/canary/.claude/skills/visual-explainer/SKILL.mdgit clone --depth 1 https://github.com/vercel-labs/vgpuWrote 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/vercel-labs/vgpu/visual-explainer)<a href="https://agentmods.dev/skills/vercel-labs/vgpu/visual-explainer"><img src="https://agentmods.dev/badge/skills/vercel-labs/vgpu/visual-explainer/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/vercel-labs/vgpu/visual-explainer"><img src="https://agentmods.dev/badge/skills/vercel-labs/vgpu/visual-explainer.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 | $0.00085 | $0.02011 |
| Opus 5.5 | $0.00034 | $0.00804 |
| Sonnet 5.5 | $0.00017 | $0.00402 |
| Haiku 4.5 | $0.00009 | $0.00201 |
Grade A, and why
visual-explainer 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 — 142 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Visual explainer
Explain the mechanism so a reader unfamiliar with the implementation can follow it from their own code to the outcome. This is the shared visual and narrative style for the repository; it does not grant permission to implement a fix or publish a PR. Follow the active workflow in AGENTS.md. Write in the user's language for personal explanations and the repository's language for PR descriptions unless directed otherwise.
Establish what is being explained
Read the issue or request, relevant source/API docs, and available reproduction or validation evidence. For a PR, identify its base and head and inspect both versions where behavior changed. Distinguish the reported symptom, confirmed cause, proposed solution and implemented result. If the fix is only proposed, label it as proposed throughout; do not present predictions as results. Use one representative example throughout the explanation. Name omitted setup and define terms when the reader first needs them. Do not infer an API from a diagram or an old draft.
Narrative: problem first, then solution
- Make the problem explicit before the first snippet. Name who encounters it, the trigger, expected behavior and actual outcome. Use a heading such as "The problem — before this PR" and state that the following steps trace the old behavior. For an unfixed issue, say "The current problem". "Summary" or "User code" alone does not establish what is being shown.
- Start at user code or an action. Show a small, representative snippet with syntax highlighting. Explain whether it is valid code exposing a library problem, incorrect usage, or a workaround. If the code is valid, do not imply the user needs to repair it.
- Follow the system step by step. Connect the user call to the necessary internal work, the failure point and its observable consequence. Use short sections, one causal step each. Separate correct mechanisms required for safety or semantics from the actual defect.
- Mark the transition to the solution. Use a separate heading such as "The solution — after this PR" or "Proposed solution". Trace the same example through the changed steps, explain why they fix the problem, and state whether user code changes. For a feature, use "Previous limitation" and "New behavior" instead of inventing a bug. For a refactor, state that observable behavior is unchanged.
- Show the result and limits. Give supported before/after evidence with units, workload, environment and conditions such as warmup. Separate allocations, CPU time and GPU throughput; mock and native results; measured outcomes and expectations. Keep unmet targets and material limitations visible in the human explanation, not only in a technical appendix.
What ships with it
1 file 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.
- 1mo ago Changed · +20 lines · +3 tokens per session 67381817e8ed
- 1mo ago First seen · 122 lines · 82 tokens per session scan A 19fe9b563647
visual-explainer is a skill published in the GitHub repository vercel-labs/vgpu (2,620 stars, last pushed 21d ago), licensed MIT. It adds 85 tokens to every session and 2,011 once invoked, about $0.0003 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 2026-10-07.
visual-explainer compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| visual-explainervercel-labs/vgpu | Skill | 2.6k | 21d ago | MIT | A | 85 | Claude Code |
| pptx-importTHU-MAIC/OpenMAIC | Skill | 43k | yesterday | MIT | A | 125 | Claude Code, Codex |
| annotate-paper54yyyu/zotero-mcp | Skill | 5.8k | 5d ago | MIT | A | 82 | Claude Code, Codex |
| dedao-dlyann0917/dedao-dl | Skill | 918 | 9d ago | MIT | A | 134 | Claude Code, Codex |
| explorable-beamerclaesbackman/AI-research-feedback | Skill | 528 | 9d ago | MIT | A | 206 | Claude Code, Codex |
| exam-study-guideZeKaiNie/universal-examprep-skill | Skill | 367 | 10d ago | MIT | A | 113 | Claude Code, Codex |
| brief-compliance-checkflonat/flonat-research | Skill | 149 | 3d ago | MIT | A | 50 | Claude Code |