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 OutlineDriven/outline-driven-development --skill cpu-kernel-authoringgit clone --depth 1 https://github.com/OutlineDriven/outline-driven-developmentWrote 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/outlinedriven/outline-driven-development/cpu-kernel-authoring)<a href="https://agentmods.dev/skills/outlinedriven/outline-driven-development/cpu-kernel-authoring"><img src="https://agentmods.dev/badge/skills/outlinedriven/outline-driven-development/cpu-kernel-authoring/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/outlinedriven/outline-driven-development/cpu-kernel-authoring"><img src="https://agentmods.dev/badge/skills/outlinedriven/outline-driven-development/cpu-kernel-authoring.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00045 | $0.02633 |
| Opus 5 | $0.00023 | $0.01316 |
| Sonnet 5 | $0.00009 | $0.00527 |
| Haiku 4.5 | $0.00005 | $0.00263 |
Grade A, and why
cpu-kernel-authoring 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 3d 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 — 67 lines — stays where its author put it; the contents beside it link to each section on GitHub.
CPU kernel authoring
Contract
| Field | Bound contract |
|---|---|
| Trigger | A C++ CPU kernel for the Hugging Face kernels ecosystem must be written, optimized, or benchmarked with AVX2 or AVX512 intrinsics against a PyTorch baseline. |
| Authority | Reversible local. Writes C++ kernel sources, build.toml, and torch_binding.cpp under the kernel directory, a wheel under dist/, the installed kernel package in the active Python environment, and trial state under trials/<kernel_name>/ and output/. Rollback is version control for the sources, pip uninstall <package> for the package, and removal of dist/, trials/<kernel_name>/, and output/. No remote mutation. |
| Side effect | Kernel sources and build files change; a wheel is built and installed; trial directories and result records accumulate. |
| Done | The kernel passes the correctness check in scripts/benchmark_cpu.py, every trial up to max_trials has run or the speedup exceeded early_stop_speedup, and the best trial is finalized into output/ with its final measurement; or a failure class from the table below is reported with the recovery step taken. |
Inputs
- Kernel name (required): the trial-tree label, for example
my_rmsnorm. Used only bytrial_manager.py, which accepts it as a single directory name undertrials/, never a path. - Baseline file (required): a
baseline.pythat definesget_inputs()and eitherget_reference_output()or aModelclass (with optionalget_init_inputs()). It is the ground truth for correctness and the speed reference. - Operation name (required): the plain name
analyze_op.py --oplooks up, for examplerms_norm. - Input shapes (required): comma-separated shape strings for
analyze_op.py --shapes, for example"1024x4096,2048x8192". - Package and function path (required from step 5): the installed package name, for example
my_kernel, and its callable aspackage.function, for examplemy_kernel.rms_norm.benchmark_cpu.pyandcpu_profiler.pytake this path as their--op; it is not the operation name above. - Toolchain (required):
kernel-builder,pip, PyYAML (imported byscripts/config.py),numactl(used by the pinned benchmark in step 8), a C++ compiler with AVX512 support, and PyTorch.perfis required only whenperf_stat_enabledis true.
What ships with it
22 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.
- agents/openai.yaml 207 B
- references/brgemm_patterns.yaml 14 KB
- references/build_system.md 7.7 KB
- references/correctness.yaml 5.9 KB
- references/dtype_optimizations.yaml 4.1 KB
- references/huggingface-kernels-integration.md 4.5 KB
- references/implementation_reference.md 12 KB
- references/memory_patterns.yaml 3.5 KB
- references/optimization_levels.yaml 5.8 KB
- references/optimization_strategies.md 4.8 KB
- references/quantized_gemm_patterns.yaml 19 KB
- references/runtime_dispatch.yaml 8.5 KB
- references/simd_optimization_patterns.yaml 7.0 KB
- references/threading_patterns.yaml 3.0 KB
- references/workflow_details.md 8.7 KB
- scripts/analyze_op.py 11 KB runs code
- scripts/benchmark_cpu.py 17 KB runs code
- scripts/config.py 1014 B runs code
- scripts/config.yaml 641 B
- scripts/cpu_profiler.py 11 KB runs code
- scripts/trial_manager.py 18 KB runs code
- scripts/validate_cpu_kernel.py 12 KB runs code
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.
- 3d ago First seen · 67 lines · 45 tokens per session scan A 5eec63c0eef7
cpu-kernel-authoring is a skill published in the GitHub repository OutlineDriven/outline-driven-development (52 stars, last pushed 3d ago), licensed Apache-2.0. It adds 45 tokens to every session and 2,633 once invoked, about $0.0002 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 2026-09-06.
Other skills, from other repositories
ax-cpp-ai
Use when writing C++ code with axllm for named deployment profiles, generic provider clients, model selection, OpenAI-compatible calls, Responses, Gemini, Anthropic, routers, and balancers.
ax-cpp-gen
Use when writing C++ code with axllm for AxGen programs, forward calls, indexed multi-sampling, result pickers, streaming, tools, assertions, traces, usage, and output parsing.
ax-cpp-llm
Use when writing C++ code with axllm for using the generated Ax package, factory functions, package docs, examples, and API reference.
ax-cpp-signature
Use when writing C++ code with axllm for string signatures, field descriptors, JSON schema output, validation, and typed tool argument shapes.
ax-cpp-gepa
Use when writing C++ code with axllm for GEPA, Pareto tradeoffs, reflection clients, metric budgets, optimizer state, and artifacts.
cuda-kernels
Provides guidance for writing and benchmarking optimized CUDA kernels for NVIDIA GPUs (H100, A100, T4) targeting HuggingFace diffusers and transformers libraries. Kernels must be kernel-builder/ABI3-compliant: no pybind11, no setup.py, TORCHLIBRARYEXPAND bindings only. Supports models like LTX-Video, Stable Diffusion…