同时搜索技能包和技能

智能体框架通常会隐藏工具如何被注册、验证和重试。此技能指导对工具抽象、schema、内置工具和错误反馈进行结构化审查。

Stale context files mislead agents and increase maintenance risk. Gardener verifies documentation against code, repairs references, and keeps risk baselines controlled.

Code changes can miss acceptance criteria, repository rules, or security boundaries. This skill performs a read-only review and returns structured, line-anchored findings.

Turn plain-English product requirements into a reactive, type-safe Convex backend design. Get guidance for schemas, functions, authentication, storage, scheduling, real-time features, and agent workflows.

Plans often fail when tasks lack owners, dependencies, or clear starting conditions. This skill turns accepted Paperclip plans into a verified task graph that supports focused and parallel execution.

开发者在构建 Python agents 时需要可靠的 ADK 模式。本技能为 tools、callbacks、state、多 Agent 编排和 workflow 代码提供简明参考。

生产环境 ADK 智能体需要跟踪、日志和分析信号，同时不暴露提示词数据。此技能指导 Cloud Trace、提示词日志记录、BigQuery 分析以及提供商选择。

启动一个 Google ADK 智能体项目需要在模板、部署和工作流方面做出多项选择。此技能会指导基于 agents-cli 的脚手架搭建，从原型设置到增强和升级规划。

Agent 输出通常需要交互式 UI，而不是纯文本。此技能展示如何将结构化 widget 定义渲染为 React 和 Next.js 组件。

GenLayer 开发需要针对账户、网络、部署和回执执行精确的 CLI 步骤。此技能可指导 Claude、Codex 和 Claude Code 安全地完成 GenLayer CLI 工作流。

Large configuration files and repeated context increase token usage. This skill reviews Caveman reports and applies measured changes with individual approval.

长对话会让跨会话继续复杂工作变得困难。此技能会创建一份聚焦的交接摘要，并启动一个带有有用上下文的后台 Claude 代理。