CtrlK
BlogDocsLog inGet started
Tessl Logo

vectorization

Guidance for writing and reviewing SIMD / hardware-intrinsics code in dotnet/runtime. USE FOR: vectorizing a scalar algorithm, writing or reviewing code that uses Vector128/Vector256/Vector512, Vector<T>, or the platform intrinsics in System.Runtime.Intrinsics.X86/Arm/Wasm, and validating remainder handling, load/store safety, and hardware-acceleration fallbacks. DO NOT USE FOR: general performance work unrelated to SIMD (use performance-benchmark), or non-vectorized code review (use code-review).

74

Quality

93%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

SKILL.md
Quality
Evals
Security

Quality

Content

86%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A tightly written, repo-specific vectorization skill that assumes Claude's competence and concentrates on enforceable rules, checklists, and concrete APIs. Its only real gap is the absence of a complete worked code example to anchor the rules.

Suggestions

Add one short, complete worked vectorized-loop example (e.g., a Vector128<T> loop with remainder masking) so the core rules and remainder handling are illustrated as copy-paste-ready code.

Make the validate→fix→retry loop explicit in the Testing checklist (e.g., 'if a path fails under DOTNET_EnableHWIntrinsic=0, fix and re-run that toggle') to push workflow clarity to 5.

DimensionReasoningScore

Conciseness

Lean and dense throughout — it offloads general SIMD background to an external doc ('Read it first and defer to it') and restricts itself to repo-specific rules, never re-explaining what SIMD or intrinsics are; every token earns its place.

5 / 5

Actionability

Highly actionable with concrete APIs (MemoryMarshal.GetReference, LoadUnsafe, ConditionalSelect, BoundedMemory.Allocate) and copy-paste env vars (DOTNET_EnableAVX2=0, DOTNET_EnableHWIntrinsic=0), but lacks a single complete worked vectorization code block; sits at 4 rather than 5 because the 5 anchor expects copy-paste examples covering common cases.

4 / 5

Workflow Clarity

Explicitly prioritized review checklist (numbered 1-7), plus authoring/testing/benchmarking checklists with validation checkpoints (BoundedMemory OOB faulting, acceleration-toggle env vars, measure-before-keep); not 5 because explicit validate→fix→retry feedback loops are implied rather than spelled out.

4 / 5

Progressive Disclosure

Well-organized into clear sections with one-level-deep, clearly signaled references (official SIMD doc, repo coding-guidelines doc, BoundedMemory source path, build-and-test and performance-benchmark skills) and no nested reference chains; easy to navigate.

5 / 5

Total

18

/

20

Passed

Description

100%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A model description: concise, third-person, with explicit USE FOR / DO NOT FOR trigger guidance and concrete API-level terms. It cleanly distinguishes itself from sibling skills and answers both what and when.

DimensionReasoningScore

Specificity

Lists multiple concrete actions in third person — 'writing and reviewing SIMD / hardware-intrinsics code', 'vectorizing a scalar algorithm', 'validating remainder handling, load/store safety, and hardware-acceleration fallbacks' — with comprehensive coverage; not the 4 anchor because no meaningful coverage gaps remain.

5 / 5

Completeness

Explicitly answers both what ('Guidance for writing and reviewing SIMD / hardware-intrinsics code') and when ('USE FOR: ... DO NOT USE FOR: ...') with concrete trigger phrases; matches the 5 anchor exactly.

5 / 5

Trigger Term Quality

Comprehensive natural-term coverage including the actual API names a contributor would say — 'SIMD', 'Vector128/Vector256/Vector512', 'Vector<T>', 'System.Runtime.Intrinsics.X86/Arm/Wasm', 'remainder handling', 'load/store safety' — covering synonyms and concrete type names.

5 / 5

Distinctiveness Conflict Risk

Clear niche (SIMD vectorization in dotnet/runtime) with an explicit 'DO NOT USE FOR' clause routing adjacent work to performance-benchmark and code-review, minimizing conflict risk.

5 / 5

Total

20

/

20

Passed

Validation

93%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

relative_links

Relative link issues: 2 suspicious

Warning

Total

15

/

16

Passed

Repository
dotnet/runtime
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.