Quark Onnx Router

What it does

When your agent uses it

• A user describes an ONNX quantization task in plain language — such as install onnxruntime
• Analyze my .onnx model
• Choose a preset for my YOLO model
• Quantize this .onnx with XINT8/BFP16/MXFP4

Example prompts

• “install onnxruntime”
• “analyze my .onnx model”
• “choose a preset for my YOLO model”
• “/quark-onnx-router”

Workflow steps

5 steps, taken from the first numbered list in SKILL.md.

1. 1Confirm the backend. If the user mentions .onnx, quantize_static, ModelQuantizer, an ONNX Runtime execution provider, opset, or QDQ, treat…
2. 2Extract the core intent. Strip filler and figure out what the user actually needs. "I want to quantize yolov8n.onnx with XINT8 for CPU" →…
3. 3Check for L0 prerequisites. Before routing to an L1 skill, check if downstream skills need facts that are missing
4. 4Pick the smallest fit. If the user only wants to inspect a .onnx graph, route to quark-onnx-model-intake — do not start the full workflow…
5. 5Handle ambiguity honestly. If the goal is unclear, record the likely routes in open_questions and ask. Example: "I want to use Quark with…

What it can do on your machine

Read from SKILL.md and the folder at commit 313cb0b. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  No scripts in the folder and no shell commands in SKILL.md (its code samples are json).

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Quark Onnx Router loads about 2.7k tokens when it runs. Until then it costs about 145 tokens; SKILL.md has 1,189 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~145

When it runs · the whole SKILL.md, loaded when a task matches

~2.7k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from amd/Quark at commit 313cb0b, republished under its MIT licence (© amd). 1,189 words, ~2,658 tokens.

name

quark-onnx-router

description

Route Quark ONNX user goals to the correct atomic skill. Use when a user describes an ONNX quantization task in plain language — such as "install onnxruntime", "analyze my .onnx model", "choose a preset for my YOLO model", "plan ONNX PTQ", "quantize this .onnx with XINT8/BFP16/MXFP4", "validate my quantized .onnx", "debug a failed ONNX quantization", or any request that involves Quark's ONNX-to-ONNX flow. This is the ONNX-side entry point — trigger it whenever the user's intent involves Quark ONNX and the correct downstream skill is not immediately obvious.

layer

l1-atomic

primary_artifact

session_context.json

source_knowledge

docs/source/install.rst, docs/source/onnx/basic_usage_onnx.rst, docs/source/onnx/onnx_examples.rst, examples/onnx/model_support.md

{
  "user_goal": "Quantize ./models/yolov8n.onnx with XINT8 and validate against the FP32 baseline",
  "workflow": "quark-onnx-ptq-workflow",
  "constraints": {
    "backend": "onnx",
    "offline": false,
    "execution_mode": "interactive_execute"
  },
  "env_context_ref": null,
  "workspace_context_ref": null,
  "onnx_install_result_ref": null,
  "quark_install_result_ref": null,
  "model_analysis_ref": null,
  "quant_plan_ref": null,
  "open_questions": [
    "Deployment target (CPU / CUDA / ROCm / NPU CNN / NPU Transformer) not yet confirmed — affects preset gating in quark-onnx-quant-plan"
  ]
}

© amd, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in .claude/skills-impl/l1-atomic/onnx/quark-onnx-router of amd/Quark.

Open the folder on GitHubat commit 313cb0b