Agent 工作流
可复用的 Agent 工作流和编排模式
Skills 列表

twill-cloud-coding-agent
Use Twill Cloud Coding Agent to manage Twill's public v1 API workflows. Create/list/update tasks, stream and cancel jobs, manage scheduled tasks, list repositories, and export Claude teleport sessions.
twillai
meticulous-iterative-dev
使用 Meticulous 进行逐步视觉验证的迭代式前端开发循环。适用于实现多步骤前端变更时,希望在最终云端测试运行之前,在每一步捕获视觉回归和意外副作用。
alwaysmeticulous
risk-management
Risk management rules learned from competition outcomes. Use when sizing positions or setting stop-losses.
0xhubed
xiaohongshu-images
Transform markdown/HTML into styled 3:4 ratio images for Xiaohongshu
iamzifei
belt
使用 belt CLI — 运行 250+ 个 AI 应用、管理知识、搜索技能、连接 MCP 服务器。专为智能体工作流设计的 CLI 接口 — 类型化输入、模式验证,无需原始 API 调用。
belt-sh
tavily-search
Tavily AI 搜索 API - 为 AI 代理优化的搜索。用于搜索当前信息、新闻、事实或任何需要实时数据的任务。
veithly
mm2-analytics-dashboard-roblox
Murder Mystery 2 inventory tracking, analytics dashboard, and gameplay optimization toolkit for Roblox
aradotso
mm2-analytics-roblox-toolkit
Roblox Murder Mystery 2 analytics dashboard and inventory tracking toolkit with data visualization and strategy analysis
aradotso
seedance-2-5-reference-to-video
Generate reference-guided 1080p video with ByteDance Seedance 2.5 Reference to Video on RunComfy via the `runcomfy` CLI. Feed up to 9 reference images, 1-3 reference video clips, and 3 reference audio files into one call and get a 4-30 second 1080p clip with native synchronized audio, identity and style locked to your references. Documents the full input schema (images / videos / audios / aspect_ratio / duration / generate_audio), the counted-seconds billing model ($0.53 per second of reference video duration plus output duration), the 480p draft-then-deliver workflow, and when to route to Seedance 2.5 text-to-video, image-to-video, or Seedance 2.0 Pro instead. Calls `runcomfy run bytedance/seedance-2.5/reference-to-video/1080p`. Triggers on "seedance 2.5", "seedance 2.5 reference to video", "reference to video", "reference-to-video", "seedance 1080p", "ByteDance Seedance 2.5", "consistent character video", "style-locked video", or any explicit ask to generate video from reference images and clips.
genmedia-labs
wordspace
CLI for bootstrapping and managing wordspace projects
frames-engineering
commit-all
User-invoked via /commit-all only. Gathers the working tree into a single commit on the current branch, no push.
sentimony
value
该技能完全不做任何事情,使用它的智能体也是如此。
hugmouse
recoup-platform-build-os
Scaffold and run a label's self-managing "music-company OS" inside its own git repo — the full label-intelligence workspace (folders, a self-managing CLAUDE.md mirrored to AGENTS.md, a read-only doctor + never-stale janitor, compound-learning and self-improvement loops, an in-place plugin) seeded with Recoup's conventions (top-level artists/{slug}/RECOUP.md identity files, releases/{slug}/RELEASE.md) and backed by the Recoup API as system of record. Use for "set up our label OS", "build the org workspace", "turn this repo into our label brain", or onboarding a new org/label repo. Pulls the real roster from the live account (auth via recoup-platform-connect-account), onboards artists API-first via recoup-roster-add-artist, and calls the other recoup-* skills (research/content/release/song/catalog) instead of reinventing them. The single workspace/OS builder — also covers the lightweight "just mirror my roster into folders" case.
recoupable
controlnet-pose
通过 `runcomfy` CLI 在 RunComfy 上进行姿态条件生成。路由涵盖 Kling 2-6 Motion Control Pro/Standard(将参考视频的动作/分镜转移到目标角色上)、社区 Wan 2-2 Animate(带姿态条件的音频驱动角色动画)以及 Z-Image Turbo ControlNet LoRA(从 OpenPose/DWPose/canny/深度控制图像进行姿态条件图像生成)。根据视频与静态图像、风格化与照片级真实感选择正确的路由。触发词包括 "controlnet"、"control net"、"pose control"、"openpose"、"DWPose"、"transfer pose"、"motion control"、"pose driven"、"character pose"、"depth control"、"canny edge"、"use this pose",或任何明确要求基于姿态/骨架/运动/深度/canny 参考进行条件生成的请求。
doany-ai
video-outpainting
>
doany-ai
controlnet-pose
通过 `runcomfy` CLI 在 RunComfy 上进行姿态条件生成。路由包括 Kling 2-6 Motion Control Pro/Standard(将参考视频的动作/走位转移到目标角色上)、社区 Wan 2-2 Animate(带姿态条件的音频驱动角色动画)以及 Z-Image Turbo ControlNet LoRA(从 OpenPose/DWPose/canny/深度控制图像进行姿态条件图像生成)。根据视频 vs 静态图像以及风格化 vs 逼真风格选择正确的路由。触发词包括 "controlnet"、"control net"、"pose control"、"openpose"、"DWPose"、"transfer pose"、"motion control"、"pose driven"、"character pose"、"depth control"、"canny edge"、"use this pose",或任何明确要求基于姿态/骨架/运动/深度/canny 参考进行条件生成的请求。
runcomfy-com
video-outpainting
>
runcomfy-com
wordspace
CLI for bootstrapping and managing wordspace projects
frames-engineering
fusion-rules
Entrypoint for AI coding assistant rule authoring across GitHub Copilot, Cursor, and Claude Code. USE FOR: setting up rules, reviewing existing rules, scaffolding instruction files, or asking which editor format to use. DO NOT USE FOR: authoring skills (SKILL.md), agent definitions (.agent.md), or CI enforcement of rule files.
equinor
agent-pulse
使用 Agent Pulse 检查本地 AI 代理活动,涵盖 Hermes、Claude Code、Codex、DeepSeek、OpenClaw、Copilot、Aider、Qwen、OpenCode、Goose、Cursor、Antigravity 和 Amp 日志。当用户询问 AI 代理会话、令牌、工具/搜索调用、模型使用、预估成本、预算、预测、健康检查、报告、设置诊断、Web/API/指标导出或 MCP 集成时使用。
jane-o-o-o-o
paperclip-converting-plans-to-tasks
Paperclip 将计划转化为可执行任务的方式。当你在 Paperclip 公司内被要求进行计划、范围界定或分解工作时使用。关于如何将计划转化为具有正确专业、依赖关系和并行化的已分配问题的行业无关指南,以便 Paperclip 的执行者能够接手工作——它不规定计划格式。与 `paperclip` 技能配合使用,该技能涵盖编写计划文档和重新分配问题的机制。
getpaperclipai
paperclip-create-agent
在 Paperclip 中创建新代理,并支持治理感知的招聘流程。当你需要检查适配器配置选项、比较现有代理配置、起草新的代理提示/配置并提交招聘请求时使用。
getpaperclipai
indexion-agent-orient
生成并使用预编辑结构简报,使编码代理在实施前了解可能的负责人、消费界面和不安全的编辑位置。
trkbt10
clawdirect
与ClawDirect交互,这是一个面向AI代理的社交网络体验目录。使用此技能浏览目录、点赞条目或添加新站点。调用MCP工具需要ATXP认证。触发条件:浏览面向代理的网站、发现代理社交平台、对目录条目点赞/投票,或向ClawDirect提交新的代理站点。
napoleond