Claude Skills 中文技能市场——为 Claude Code、Codex 与 Cursor 精选 Agent Skills

收录 90,000+ 开源 SKILL.md 工作流,提供中文解读、安全检测与高速下载,一行命令即可装到你的 AI 编程助手里。

加载中…

高分中文技能

evaluation

Asmayaseen's avatarfrom Asmayaseen

Build evaluation frameworks for agent systems. Use when testing agent performance, validating context engineering choices, or measuring improvements over time.

ai-feedback-loop-optimizer

ntaksh42's avatarfrom ntaksh42

AI反馈循环优化技能。自动执行提示→输出→评估→改进的迭代循环。通过渐进式改进、A/B测试、收敛判断和最佳输出选择生成最高质量的结果。

evaluation-metrics

ricardoroche's avatarfrom ricardoroche

在评估LLM性能时自动应用。确保适当的评估数据集、指标计算、A/B测试、LLM作为判断模式和实验跟踪。

use-skill-create

mtthsnc's avatarfrom mtthsnc

Use when creating new skills, editing existing skills, or verifying skills work before deployment

confidence-evaluator

davjdk's avatarfrom davjdk

使用ISO/IEC/IEEE 29148:2018标准评估需求清晰度和完整性。当用户请求实现功能、修复错误或进行更改时使用。在ai-settings.json中启用confidence_policy时自动调用。

llm-call

majiayu000's avatarfrom majiayu000

外部LLM调用。仅由@council、@probe、@crossref、@gpt、@gemini、@grok、@qwen触发。

taiga-api

atondwal's avatarfrom atondwal

查询taiga.ant.dev上托管的Taiga API以获取作业结果、通过率、转录本和运行评估。当用户询问Taiga作业、问题分数、评估结果或需要提交/检查作业时使用。

agent-audit

philoserf's avatarfrom philoserf

Validates agent configurations for model selection appropriateness, tool restriction accuracy, focus area quality, and approach completeness. Use when reviewing, auditing, or improving agents, checking model choice (Sonnet/Haiku/Opus), validating tool permissions, assessing focus area specificity, or ensuring approach methodology is complete. Also triggers when user asks about agent best practices, wants to optimize agent design, or needs help with agent validation.

audit-coordinator

philoserf's avatarfrom philoserf

Orchestrates comprehensive audits across multiple specialized auditors for Claude Code customizations. Use when user wants complete evaluation of agents, skills, hooks, commands, output-styles, entire setup, needs multi-faceted analysis, wants coordinated audit reports, requests thorough validation, or asks to audit multiple components. Automatically determines which auditors to invoke (agent-audit, skill-audit, hook-audit, command-audit, output-style-audit, evaluator, test-runner) based on target type and compiles unified reports with consolidated recommendations.

output-style-audit

philoserf's avatarfrom philoserf

Validates output-style persona definitions, behavior specifications, and keep-coding-instructions decisions. Use when reviewing, auditing, or improving output-styles, checking persona clarity, validating behavior concreteness, assessing coding-instructions appropriateness, or verifying scope alignment (user vs project). Also triggers when user asks about output-style best practices, wants to improve style effectiveness, or needs help with persona definition.

skill-audit

philoserf's avatarfrom philoserf

Audits skills for discoverability and triggering effectiveness. Use when analyzing skill descriptions, checking trigger phrase coverage, validating progressive disclosure, reviewing SKILL.md structure, ensuring skills are discoverable, testing skill triggering, improving skill descriptions, or evaluating whether a skill will be invoked appropriately. Also triggers when user asks about skill best practices, wants to improve skill discovery, or needs help with skill structure.

checking-skill-best-practices

onukura's avatarfrom onukura

评估Claude技能是否符合Anthropic官方最佳实践文档。在审查技能质量、确保符合指南或改进现有技能时使用。

llm-eval-designer

CAPHTECH's avatarfrom CAPHTECH

LLM生成システムの検証設計スキル。assay-kitフレームワークを活用し、LLM特有の失敗モード(幻覚、例への過学習、部分的処理)を考慮した総合的なテストケース設計を支援する。使用タイミング:- LLMベースのワークフロー/エージェントの評価設計時- ゴールデンデータセット(golden-dataset.yaml)の設計・拡張時- 既存テストが特定パターンに過学習していないか検証時- LLM出力の品質スコアラー設計時- 「なぜこのテストケースが必要か」の根拠を示す時

embedding-comparison

mindmorass's avatarfrom mindmorass

比较和评估语义搜索的嵌入模型。

base-model-selector

marcgreen's avatarfrom marcgreen

Use when starting a fine-tuning project to determine if fine-tuning is needed, or when evaluating whether a base model meets quality thresholds for a specific domain task

finetune-design

marcgreen's avatarfrom marcgreen

在准备微调LLM以进行多轮对话之前,用于生成任何训练数据时使用。触发条件:开始微调项目,需要定义评估标准,设计对话数据生成。

finetune-train

marcgreen's avatarfrom marcgreen

在训练微调模型并评估相对于基础模型的改进时使用。触发条件:已过滤训练数据,准备提交训练任务,需要转换为GGUF。需先使用finetune-generate。

finetune-generate

marcgreen's avatarfrom marcgreen

用于生成多轮对话微调的合成训练数据。触发条件:设计工件准备就绪,需要生成对话,准备评估质量。需先进行微调设计。

常见问题

什么是 Claude Skills(Agent Skills)?

Skills 是一种标准化的 SKILL.md 指令包——把「何时触发、如何执行、使用哪些工具」写成结构化工作流,装入 Claude Code、Codex、Cursor 等 AI 编程助手后,Agent 遇到对应任务会自动调用,效果远好于每次手写提示词。

装了 Skill 之后怎么使用?

安装后无需手动调用:直接在 Claude Code / Cursor 里描述你的任务(如「帮我做 PR 代码审查」),Agent 检测到匹配的技能后会自动加载对应 SKILL.md 并按其流程执行。也可以在对话中显式要求「使用 xx skill」。

国内如何高速下载安装?

本站在 Cloudflare R2 托管了全部技能 ZIP 离线包,无需直连 GitHub:在详情页点击「Download ZIP」,或用 `npx skills add <owner/repo>` 一键安装;也可以复制 curl 命令直接解压到 ~/.claude/skills 目录。