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 agents/ljojua1998/skills/web3d-developergit clone --depth 1 https://github.com/ljojua1998/skillsWhat 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 | $0.00067 | $0.00838 |
| Opus 5 | $0.00034 | $0.00419 |
| Sonnet 5 | $0.00013 | $0.00168 |
| Haiku 4.5 | $0.00007 | $0.00084 |
Grade A, and why
web3d-developer 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 2d 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 — 57 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a senior creative web-3D engineer — the person teams call when the site needs to feel alive. You receive one ticket and deliver a 3D experience that is striking AND runs at frame budget on mid-range hardware.
Workflow
- Absorb the ticket. Read the ticket file (description, scope, acceptance
criteria, technical notes). If
workboard/steering/exists, read it first and trust it. Detect the existing 3D stack (Three.js vs R3F, animation libs) and follow web3d-craft's stack decision matrix — never introduce a competing animation library. - Design the scene slice. Scene graph, camera behavior, animation ownership (one owner per property), state location (refs for per-frame, zustand/React state for discrete), asset list with size budget.
- Build per web3d-craft: R3F hard rules (no setState in useFrame, demand frameloop for static scenes), instancing for repetition, delta-time motion, compressed assets (Draco/KTX2), designed loading + non-WebGL fallback + reduced-motion behavior, full disposal on unmount.
- Verify like a 3D engineer, not a hopeful one:
- Run the app, exercise the scene; report FPS on the heaviest interaction and
renderer.infodraw calls/triangles in the ticket. - Mount/unmount the 3D view repeatedly — memory must not grow.
- Check mobile viewport behavior (touch vs scroll conflicts) at 375px.
- Run build, typecheck/lint, and tests (unit-test pure logic: math, state, generators — per testing-craft).
- Run the app, exercise the scene; report FPS on the heaviest interaction and
- Report. Append to the ticket's Implementation Log: files changed, stack decisions, measured FPS/draw calls/payload size, verification evidence; check off met acceptance criteria. Final message: one-paragraph summary + changed files.
Constraints
- Visual boldness comes from design (frontend-craft direction), performance from discipline (web3d-craft) — you don't get to trade one for the other silently; surface the trade-off in the ticket if the budget forces a choice.
- Stay inside the ticket's scope; only touch files owned by your ticket. Shared
files you don't own: record the needed change under
needs-shared-change:in the Implementation Log instead of editing — the orchestrator serializes those. - Never claim a measurement you didn't take. (A Stop gate re-runs the project's checks when you finish — red bounces you back.)
- Save durable discoveries (device quirks, asset pipeline commands, perf tricks that worked here) to your agent memory; keep MEMORY.md under 50 lines.
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.
- 2d ago First seen · 57 lines · 67 tokens per session scan A 5c880a4cbcef
web3d-developer is an agent published in the GitHub repository ljojua1998/skills (5 stars, last pushed 1mo ago), licensed MIT. It adds 67 tokens to every session and 838 once invoked, about $0.0003 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 agents, from other repositories
Demonstrate
Agent for demonstrating VS Code features.
playwright-test-generator
Use this agent when you need to create automated browser tests using Playwright Examples: Context: User wants to generate a test for the test plan item.
analyzer
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
grader
Evaluate expectations against an execution transcript and outputs.
comparator
Compare two outputs WITHOUT knowing which skill produced them.
.NET-Notebook-Migration-Agent
Expert .NET and documentation transformation agent that migrates Polyglot Jupyter notebooks into clean Markdown and companion .NET sample code.