ns-stdlib — compiler ns-stdlib, neuroscript-rs, community, compiler, ide skills, deep-learning, domain-specific-language, neural-networks, neuroscript, pytorch, Claude Code

v1.0.0
GitHub

关于此技能

适合需要复合神经架构和NeuroScript标准库集成的高级AI代理。 NeuroScript standard library catalog. Lists all primitive and composite neurons with signatures, shapes, parameters, and categories. Use when looking up available neurons, checking signatures, or finding which neuron to use.

# 核心主题

compiler deep-learning domain-specific-language neural-networks neuroscript pytorch
severeon severeon
[0]
[0]
更新于: 3/12/2026

Killer-Skills Review

Decision support comes first. Repository text comes second.

Reference-Only Page Review Score: 6/11

This page remains useful for operators, but Killer-Skills treats it as reference material instead of a primary organic landing page.

Original recommendation layer Concrete use-case guidance Explicit limitations and caution
Review Score
6/11
Quality Score
27
Canonical Locale
en
Detected Body Locale
en

适合需要复合神经架构和NeuroScript标准库集成的高级AI代理。 NeuroScript standard library catalog. Lists all primitive and composite neurons with signatures, shapes, parameters, and categories. Use when looking up available neurons, checking signatures, or finding which neuron to use.

核心价值

赋予代理使用NeuroScript原语（如FFN、残差、MultiHeadAttention和TransformerBlock）组合复杂神经网络的能力，利用NeuroScript标准库目录实现高级神经架构。它可以创建自定义的复合库神经元，并提供类别索引以实现高效导航。使用grep、sed和NeuroScript的列表功能实现神经组件的动态注册和检查。

适用 Agent 类型

适合需要复合神经架构和NeuroScript标准库集成的高级AI代理。

↓ 赋予的主要能力 · ns-stdlib

使用NeuroScript原语组合自定义神经网络
注册和检查复合库神经元
利用NeuroScript标准库目录进行神经架构设计

! 使用限制与门槛

  • 需要NeuroScript的安装和配置
  • 依赖于特定的NeuroScript原语和复合库神经元
  • 仅限于NeuroScript兼容的神经架构

Why this page is reference-only

  • - Current locale does not satisfy the locale-governance contract.
  • - The underlying skill quality score is below the review floor.

Source Boundary

The section below is supporting source material from the upstream repository. Use the Killer-Skills review above as the primary decision layer.

实验室 Demo

Browser Sandbox Environment

⚡️ Ready to unleash?

Experience this Agent in a zero-setup browser environment powered by WebContainers. No installation required.

Boot Container Sandbox

常见问题与安装步骤

以下问题与步骤与页面结构化数据保持一致，便于搜索引擎理解页面内容。

? FAQ

ns-stdlib 是什么？

适合需要复合神经架构和NeuroScript标准库集成的高级AI代理。 NeuroScript standard library catalog. Lists all primitive and composite neurons with signatures, shapes, parameters, and categories. Use when looking up available neurons, checking signatures, or finding which neuron to use.

如何安装 ns-stdlib？

运行命令：npx killer-skills add severeon/neuroscript-rs。支持 Cursor、Windsurf、VS Code、Claude Code 等 19+ IDE/Agent。

ns-stdlib 适用于哪些场景？

典型场景包括：使用NeuroScript原语组合自定义神经网络、注册和检查复合库神经元、利用NeuroScript标准库目录进行神经架构设计。

ns-stdlib 支持哪些 IDE 或 Agent？

该技能兼容 Cursor, Windsurf, VS Code, Trae, Claude Code, OpenClaw, Aider, Codex, OpenCode, Goose, Cline, Roo Code, Kiro, Augment Code, Continue, GitHub Copilot, Sourcegraph Cody, and Amazon Q Developer。可使用 Killer-Skills CLI 一条命令通用安装。

ns-stdlib 有哪些限制？

需要NeuroScript的安装和配置；依赖于特定的NeuroScript原语和复合库神经元；仅限于NeuroScript兼容的神经架构。

↓ 安装步骤

  1. 1. 打开终端

    在你的项目目录中打开终端或命令行。

  2. 2. 执行安装命令

    运行：npx killer-skills add severeon/neuroscript-rs。CLI 会自动识别 IDE 或 AI Agent 并完成配置。

  3. 3. 开始使用技能

    ns-stdlib 已启用，可立即在当前项目中调用。

! 参考页模式

此页面仍可作为安装与查阅参考，但 Killer-Skills 不再把它视为主要可索引落地页。请优先阅读上方评审结论，再决定是否继续查看上游仓库说明。

Imported Repository Instructions

The section below is supporting source material from the upstream repository. Use the Killer-Skills review above as the primary decision layer.

Supporting Evidence

ns-stdlib

安装 ns-stdlib，这是一款面向AI agent workflows and automation的 AI Agent Skill。支持 Claude Code、Cursor、Windsurf，一键安装。

SKILL.md
Readonly
Imported Repository Instructions
The section below is supporting source material from the upstream repository. Use the Killer-Skills review above as the primary decision layer.
Supporting Evidence

NeuroScript Standard Library Catalog

Available Primitives (live)

!grep -A1 'self\.register(' src/stdlib_registry.rs | grep '"' | sed 's/.*"\([^"]*\)".*/\1/' | sort

Composite Library Neurons (live)

!for f in stdlib/FFN.ns stdlib/Residual.ns stdlib/MultiHeadAttention.ns stdlib/TransformerBlock.ns stdlib/TransformerStack.ns stdlib/MetaNeurons.ns; do [ -f "$f" ] && echo "=== $f ===" && ./target/release/neuroscript list "$f" 2>/dev/null; done

Category Index

CategoryPrimitivesUse For
CoreLinear, Bias, Scale, MatMul, EinsumDense layers, linear transforms
ActivationsGELU, ReLU, Tanh, Sigmoid, SiLU, Softmax, Mish, PReLU, ELUNon-linearities
NormalizationLayerNorm, RMSNorm, GroupNorm, BatchNorm, InstanceNormStabilizing training
RegularizationDropout, DropPath, DropConnectPreventing overfitting
ConvolutionConv1d, Conv2d, Conv3d, DepthwiseConv, SeparableConv, TransposedConvSpatial feature extraction
PoolingMaxPool, AvgPool, AdaptiveAvgPool, GlobalAvgPool, AdaptiveMaxPool, GlobalMaxPoolSpatial reduction
EmbeddingsEmbedding, PositionalEncoding, LearnedPositionalEmbedding, RotaryEmbeddingToken/position encoding
StructuralIdentity, Fork, Fork3, ForkN, Add, Multiply, Concat, Reshape, Transpose, Flatten, Split, Slice, PadRouting and reshaping (implicit fork preferred for splitting)
AttentionScaledDotProductAttention, MultiHeadSelfAttentionAttention mechanisms
DebugLogDebugging tensor flow

Decision Tree: Which Neuron?

Need to transform features? → Linear(in_dim, out_dim) Need non-linearity? → GELU() (default), ReLU() (legacy), SiLU() (modern) Need normalization? → LayerNorm(dim) (transformer), RMSNorm(dim) (efficient), BatchNorm(dim) (CNN) Need residual connection? → in -> (main, skip) + processing + Add() (implicit fork) Need N-way split? → in -> (a, b, c, ...) (implicit fork — any number of outputs) Need to concatenate? → Concat(dim=-1) — takes 2 inputs via named ports Need attention? → MultiHeadSelfAttention(d_model, heads) (complete) or compose from ScaledDotProductAttention(d_k) Need convolution? → Conv2d(in_ch, out_ch, kernel) (standard), SeparableConv(...) (efficient) Need position info? → PositionalEncoding(seq, dim) (sinusoidal), RotaryEmbedding(dim, seq) (modern)

Standard Library Composites

The stdlib/ directory provides higher-level neurons built from primitives:

  • FFN.ns — Feed-forward networks: FFN(dim, expansion), FFNWithHidden(in, hidden, out)
  • TransformerBlock.ns — SimpleTransformerBlock(dim), TransformerBlock(dim, heads, d_ff)
  • TransformerStack.ns — TransformerStack2(d, heads, d_ff), SequentialTransformer(d, heads, d_ff)
  • MetaNeurons.ns — ParallelFFN(dim) and routing patterns

See references/primitives-by-category.md for full signatures. See references/composite-library.md for stdlib neuron details. See references/impl-format.md for how impl references map to Python.

相关技能

寻找 ns-stdlib 的替代方案 (Alternative) 或可搭配使用的同类 community Skill？探索以下相关开源技能。

查看全部

openclaw-release-maintainer

Logo of openclaw
openclaw

Your own personal AI assistant. Any OS. Any Platform. The lobster way. 🦞

★ 333.8k
⑂ 0
AI

widget-generator

Logo of f
f

为prompts.chat的信息反馈系统生成可定制的插件小部件

★ 149.6k
⑂ 0
AI

flags

Logo of vercel
vercel

React 框架

★ 138.4k
⑂ 0
浏览器

pr-review

Logo of pytorch
pytorch

Python中具有强大GPU加速的张量和动态神经网络

★ 98.6k
⑂ 0
开发者工具