YAO = Yielding AI Outcomes. A rigorous engineering, evaluation, governance, and portability system for reusable agent skills.
-
Updated
Nov 9, 2026 - Python
YAO = Yielding AI Outcomes. A rigorous engineering, evaluation, governance, and portability system for reusable agent skills.
From thought to skill. From signal to structure.
12 evidence-backed skill-writing principles plus repository engineering for real-world agent skills, inspired by Matt Pocock / 十二条有证据支撑的 Skill 写作原则与仓库工程指南
Skill Creator Pro 是一款以行为为核心的设计工具,完全基于 Codex 框架开发。该工具可用于设计、构建、验证和发布新的智能体技能。它扩展了 OpenAI 的 Skill Creator 功能,增加了明确的行为契约、可观察的工作流程机制、收集级架构设计、质量检查功能以及实用的前向测试机制。
Turn goals, existing Skills, and authorized work records into bounded Agent Loops, then export a Codex Skill or Compact Prompt with separate verifiable evidence.
Audit any Claude Code SKILL.md file. Scored report with critical issues, triggering score, and exact fixes. Free skill-engineering tool by Anvil & Code.
Production-grade Claude Code skill engineering tools. Scaffolder, validator, security auditor, tester, optimizer and more. Build once, ship reliably.
Optimize Claude Code SKILL.md description fields for maximum trigger reliability. 3 scored variants, clear recommendation. Free skill-engineering tool by Anvil & Code.
三线 Agent Skill 体系——47 个原子技能。这不是一份技能清单,而是一套方法论的沉淀:让 AI 助手在「软件开发生命周期」、「知识管理」与「陌生领域学习」三条工作流上,每次都以相同的过程工作。
Meta-skills for creating, reviewing, and maintaining production-quality Agent Skills.
Edge AIOps — Multi-Agent Incident Response for Edge + AI Fleets (AgentTeams). Demo for Agent Infra competition.
AI Skill 的开发工作台与工厂机制(冻结思想展示版)——开发态/发布态分离、配置驱动多 Agent 分发、四个元 skill 方法论。
多智能体拆分判定与编排治理技能 — 三问耦合体检/单Agent基线任务集/交接契约五字段/证据审查/同任务复测四数字(拆不拆由任务耦合度决定,默认不拆)
Skill Engineering 实践课程:课件、四步跟练、进阶练习与自评工具。
Skill Engineering Stack v5.5 — A multi-pass compiler that transforms natural-language sources into formal, typed, executable generative models
开源 Skill 库的「事实核对 + 三层迭代」方法论 — N1-N4 核对(Star 数是曲线不是快照/宣传清单≠全量/模拟护栏读原文/评测限制=必做项)→底座先行(product-marketing.md 四要素+替换测试)→先审计后战略(没数据跑 council=空转)→定制四动作(中文 AI 本土化/顾问团两级标准/1-2 个 loop 跑通闭环/veto list 进配置)→补它没做的(每个 Skill 配基线+指标+回归测试 + 分歧地图机制抽到定价/产品路线/技术选型)(源自 marketingskills 47 Skill 拆解)
7-Aspects for AI Agent Skill Engineering — VDA 6.3 Process Turtle Diagram framework for systematic Skill design and evaluation (CN/EN)
To associate your repository with the skill-engineering topic, visit your repo's landing page and select "manage topics."