查找技能

通过grizzly在macOS的Bear中创建、读取和追加笔记，按标签查找内容。还可获取JSON格式的笔记内容或标签列表，并在执行前预览操作URL。

• 创建带标签的工作笔记
• 向已有笔记追加信息
• 按标签查找笔记

通过Obsidian官方CLI搜索、创建和编辑笔记，更新每日记录与任务，检查链接和属性，也可指定操作的笔记库。

• 查找会议笔记与项目记录
• 向每日笔记追加记录
• 查看并完成未完成任务

通过Mac查看、添加和编辑Apple Reminders中带截止时间的待办事项，管理列表并完成或删除任务，整理在iPhone或iPad上查看的个人任务。

• 查看今天和已逾期的任务
• 添加带截止日期的个人待办事项
• 整理iPhone或iPad上的任务

通过代理管理Trello中的看板、列表和卡片。可查看任务及卡片详情，创建卡片、在列表间移动卡片、添加评论，并归档已完成的卡片。

• 查看看板中的任务
• 将新任务创建为卡片
• 按进度移动卡片

通过gog让智能体处理Google Workspace日常工作，包括搜索和发送Gmail邮件、管理Calendar日程、查找Drive文件，以及读写Sheets表格。

• 搜索近期邮件并起草回复
• 创建和更新日程
• 查找文件并查看联系人

从CodexBar本地费用日志中，按模型汇总Codex或Claude的使用费用，以文本或JSON查看最近的代表模型、全部模型或指定模型的费用。

• 查看最近代表模型的费用
• 比较全部模型的使用费用
• 汇总已导出的费用日志

规定会话开始时检查相关技能，并在回复或执行任务前加载的流程。优先使用决定工作方法的技能，再衔接实施技能及各运行环境的使用指南。

• 在会话开始时检查技能
• 在开发功能前选择技能
• 在调查故障前选择技能

Ask the agent to explain its last message again with some context. The revised explanation uses ASD-STE100, a simplified technical English standard, and the vocabulary defined in the project's glossary.

中文简介正在准备中。当前显示英语简介。

• Requesting a clearer explanation of the last message
• Adding context when the discussion becomes hard to follow
• Explaining with the project's agreed terminology

Refine a plan or design through a persistent interview that asks probing questions. This skill starts the conversation by invoking the related grilling skill, providing an entry point for examining and sharpening what you have in mind.

中文简介正在准备中。当前显示英语简介。

• Refining a plan through questions
• Examining a design in an interview
• Explaining an initial idea through dialogue

将进行中的对话整理为交接摘要，移除敏感信息，明确后续重点、参考资料和建议技能，再启动新的后台智能体，从当前工作目录继续处理任务。

• 将后续工作交给新的智能体
• 限定下一次会话的任务范围
• 整理下一次会话所需的参考资料

通过附有建议答案的问题，梳理工作和生活中的重复任务，明确启动条件、人工确认环节及决策摘要，形成实施人员无需追加提问即可据此构建的工作流规格。

• 寻找可以委托的重复任务
• 明确工作流的启动条件
• 确定人工确认的时机

Summarize the current conversation into a handoff document so another agent can continue the work. Reference existing project materials, suggest skills for the next session, and redact sensitive information.

中文简介正在准备中。当前显示英语简介。

• Handing ongoing work to another agent
• Preparing context for a new session
• Creating a handoff focused on the next task

在单次会话中协调承担不同任务的子智能体，推进长期目标。后台工作与用户对话同时进行，并改进共享规范、工作环境和后续任务所需的信息访问条件。

• 协调长期目标中的任务分工
• 后台推进工作并持续对话
• 改进共享规范和工作环境

Review Claude skills and commands for usefulness, overlapping content, and outdated technical references. Choose a scan of changed skills or a full review, then receive reasoned recommendations to keep, improve, update, retire, or merge them.

中文简介正在准备中。当前显示英语简介。

• Reviewing all installed skills
• Checking only recently changed skills
• Finding overlapping skill content

Browse available specialist agents by domain, select up to five, and run them independently on the same task in parallel. Receive a combined report showing each agent’s findings, shared conclusions, conflicting recommendations, and suggested next steps.

中文简介正在准备中。当前显示英语简介。

• Browsing available specialist agents
• Selecting agents from different domains
• Reviewing a site for security and SEO before launch

Find existing skills before building a new one. Search local installations, marketplaces, GitHub, and the web, then review candidates for relevance and risky behavior to help choose between reuse, extension, and fresh creation.

中文简介正在准备中。当前显示英语简介。

• Finding a skill for a specific workflow
• Checking for duplicates before building a skill
• Choosing an existing skill to extend

Prepare and inspect plans for opening interactive commands in a visible WezTerm window while keeping the executable and arguments separate. Check terminal availability, preview launches as JSON, and use fallback or recovery modes when needed.

中文简介正在准备中。当前显示英语简介。

• Opening an interactive CLI in a new terminal window
• Starting an SSH session or local development process
• Inspecting a launch plan without opening a window

Help manage long Claude Code sessions by suggesting when to summarize accumulated context with /compact. Uses context size and tool calls as signals, with guidance on saving plans and choosing suitable breaks between task phases.

中文简介正在准备中。当前显示英语简介。

• Managing long sessions near the context limit
• Summarizing research before implementation
• Switching tasks after debugging

Choose how detailed an answer should be before it is written, using four levels and estimated token counts, a measure of text length. The skill then adjusts the response and keeps the chosen level for subsequent replies.

中文简介正在准备中。当前显示英语简介。

• Choosing between a brief answer and a detailed explanation
• Checking estimated token use before an answer
• Requesting a balanced answer with one example

Turn a multi-step plan into ready-to-paste ECC slash commands, text commands entered with a slash. Each generated command includes its step identifier, task scope, and acceptance criteria. Steps that cannot carry their scope are marked as blocked for manual handling.

中文简介正在准备中。当前显示英语简介。

• Turning an implementation plan into commands for each step
• Choosing commands for feature additions, fixes, and refactoring
• Generating commands for a selected step or range

Present local Markdown plans and HTML artifacts in a browser for review. People can annotate specific elements, chat, and approve or request changes, while the agent receives structured feedback and file edits appear automatically.

中文简介正在准备中。当前显示英语简介。

• Reviewing and approving an implementation plan
• Pointing to specific changes in a design
• Annotating a report or comparison

Organize a task into parallel work streams with clear dependencies, separate places for changes, and explicit checks. Batch independent reads and tests, coordinate concurrent work, and report completed or blocked tasks with verification results.

中文简介正在准备中。当前显示英语简介。

• Batching independent file reads and searches
• Splitting implementation across multiple agents
• Separating changes with Git worktrees

Set up human approval for agent-drafted messages to customers or partners. Bind decisions to exact text and destinations, record delivery evidence, and prevent outdated approvals or competing workers from triggering unintended sends.

中文简介正在准备中。当前显示英语简介。

• Require approval before sending customer replies
• Keep an audit trail of approved text and delivery
• Prevent old approvals from releasing revised drafts

Retrieve messages from iMessage or social DMs, find recent one-time codes, and inspect conversations before preparing replies. The workflow identifies the source, sender, and time window, then reports what was found or the specific access blocker.

中文简介正在准备中。当前显示英语简介。

• Checking recent texts or DMs
• Finding a recent one-time code
• Reading a conversation before preparing a reply