Best AI Agent AI SkillsPage 7

OpenClaw Godmode: Orchestrates multi-agent development workflows from a user-stated goal.

Accesses the Google Tasks API through managed OAuth to manage task lists and tasks.

A self-improvement workflow covering research, meta-learning, and constrained changes

个人数字工作台搭建专家，零基础小白也能用。用户不用写长提示词——只说一句场景（如「做个记账台」），技能自动匹配内置专家预设并生成单文件 HTML 工作台：手机电脑通用、本地存储不丢、可一键部署到手机桌面当 APP。覆盖记账/减脂/备考/育儿启蒙/自媒体/学生/个人IP/生活管理等场景。触发词「工作台」「搭建」「记账」「打卡」「待办」「备考」「启蒙」「识字」「内容创作」「做个XX台」「想要一个网页」等。

Records and retrieves long-term memory locally with bash and jq.

提示词优化工具，将模糊的需求描述转化为精确、结构化的 AI 提示词。基于 EARS 方法论 （易用需求规格语法），帮助写出更有效的 prompt。适用于优化 AI 写作提示、代码生成需求、 产品功能描述等场景。解决"prompt 效果不好"、"AI 理解不了我的需求"、"怎么写好提示词"、 "提示词优化 / prompt 工程"等问题。

Analyze recent conversation corrections, preferences, and lessons to help update an agent’s persistent memory, configuration, and learning logs.

Organizes story creation, chapter review and multilingual translation projects.

Provides comprehensive guidance for crafting image-to-image (图生图) edit prompts for 即梦 (Dreamina/Jimeng) models 4.0+. Uses a Keep/Change framework: explicitly declare what stays and what transforms. Use when the user wants to modify an existing image via style transfer (风格迁移), background replacement (背景替换), detail enhancement (细节增强), color adjustment (色彩调整), element add/remove (元素添加移除), season/time change (季节时间变换), restoration (修复), outfit change (换装), creative transformation (创意变形), or multi-reference compositing (多图合成); mentions \"图生图\", \"image2image\", \"编辑图片\", \"换背景\", \"改风格\", \"修图\", \"把这张图变成\"; or provides a reference image and asks to transform it. Covers 10 edit categories with 30 annotated examples and I2I-specific Keep/Change templates. Always use this skill when the user wants to edit an existing image using 即梦.

Recovers the minimum conversation history needed to resume after compaction or truncation.

Semi-automatically replies to WeChat messages and asks you when confidence is low.

Orchestrates the skills installed in OpenClaw, resolving dependencies and scheduling.

Create or upgrade OpenClaw skills with a first-use setup wizard and reusable scaffolding.

Suggest contextual next questions in OpenClaw conversations.

A fast Rust-based headless browser automation CLI with Node.js fallback that enables AI agents to navigate, click, type, and snapshot pages via structured co...

基于100+验证模板生成高质量AI提示词。当用户想要创建、优化或改进ChatGPT、Claude等AI工具的提示词时使用此技能。自动选择最佳模板类别，填充占位符，提供可直接使用的提示词。支持写作、绘图、编程、学习、办公等多种场景。

A framework for weather-market strategies on Polymarket that defaults to dry-run mode.

Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express...

Creates Docker Sandbox environments with outbound proxy controls.

Organizes and retrieves WorkBuddy memories for review before starting a task.

lancedb-memory provides code for local long-term memory storage and retrieval with LanceDB.

Manages Tencent Cloud object storage and its media processing service.

发送“领取”即可自动领取 WorkBuddy 积分（Buddy 加油站每日签到 + 派猫猫旅行先领后派）。纯标准库、幂等安全、不打印 token，适合每日定时自动化。

房地产经纪人客户需求分析技能（花桥学堂BL1课程方法论）。当用户上传客户沟通信息（微信聊天记录、电话记录、带看反馈、客户基本情况等）并希望分析购房客户需求、生成《客户需求分析报告》时使用此技能。基于"人生六套房+需求冰山+5W2H+SPIN+人生叙事"模型，从显性需求深入隐性需求和深层需求，产出可直接发送给客户阅读的HTML报告。触发词：客户需求分析、需求解码、分析客户、生成需求报告、懂客户。

fluid-memory is an anthropomorphic fluid-memory system that fully relies on OpenClaw's native memory system.

Records errors, corrections and lessons as traceable project notes.

A WeChat messaging skill that combines window control with screen-based visual understanding.

Lets an agent remember facts, lessons and entities across sessions.

腾讯营销（原腾讯广告）技能鉴权凭证管理技能，负责检查 APIKEY 是否有效，当腾讯营销(tencent-ads)系列技能返回认证错误时引导用户提供 APIKEY 并保存(⚠️重要提醒⚠️：腾讯营销 APIKEY 必须通过该技能保存)，所有 tencent-ads 技能共享。

写作团队Lead管理专业写手,分析写作任务并分发,通过协议工具协调多人创作

Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.

An encrypted messaging skill for private agent-to-agent communication.

按任务场景，输出可复用的高质量提示词模板，含结构化框架与最佳实践。适用场景： AI 应用开发者/产品经理/内容创作者快速产出高质量注意：本技能依据用户给定资料生成结构化方案，仅尽力提升质量，不承诺具体效果，最终决策与落地需用户自行复核。

创建、优化、审计、安全门禁和打包 AI Skill 技能包；支持新手引导模式（第一次做 Skill、一步步教我、从零创建），覆盖提示注入、硬编码凭据、危险命令、动态构造命令、反序列化、依赖安装、供应链、路径越界、外联声明与敏感文件名等 15 项安全检测并如实披露检测边界。当用户要求创建技能、生成 Skill、优化 SKILL.md、修复技能包结构、补齐 YAML front matter、做静态校验与安全审计、运行 Gate 判定、打包发布 SkillHub，或把业务流程沉淀为可安装技能时使用。不要用于一次性提示词润色、普通文档写作或与 Skill 无关的软件开发任务。

Windows 电脑端微信消息监控与发送 MCP，实现在微信上给指定联系人发送消息

Start or close the Douyin desktop client through scripts.

memory-qdrant documents semantic conversation memory with local embeddings.

Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Clau...

Search ClawHub skills or list locally installed skills.

A marketplace integration skill that connects human tasks with AI agents.

A skill that calls Baidu Map web service APIs.

A proactive goal and task management system skill for agents.

Covers context compression and summarization strategy, aimed at total cost per task.

A skill that compresses the context of an OpenClaw session.

Self-Improving helps an agent learn from work through reflection, correction, and tiered memory.

reflection teaches agents to pause and self-review: self-critique before delivery, learn from corrections and detect repeating patterns, cutting revision rounds.

信息检索和搜索领域中的网页搜索功能。当用户想要 AI 优化的搜索接口、专为 Agent 场景设计且返回简洁相关结果时使用。触发器：tavily-search、gemini、valyu。不适用于与网页搜索无关的请求。风险等级 L0。中文触发：tavily-search、gemini、valyu。本技能为 SkillHub 按次计费技能（0.01 元/次）：执行核心任务会触发一次平台支付授权，是否执行付费动作由用户决定；授权由作者自有服务 pay.sharpdealhub.com 承载（非第三方支付网关），仅接收动作名与订单号，不读取本技能处理的内容。

Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express...

End-of-session knowledge cleanup with OCD-level rigor — reconciles project docs (CLAUDE.md, README.md, docs/) and agent memory against the code, and audits w...