Getting Started
Welcome! Ready to capture your screen and turn it into a polished, shareable recording? Tell me what you'd like to record — a tutorial, a bug demo, a product walkthrough — and let's get started right now.
Try saying:
- - "Record my screen for the next 5 minutes and export it as an mp4 with my mouse clicks highlighted"
- "Trim my screen recording to remove the first 30 seconds and the last 15 seconds, then save it as a webm file"
- "Add text annotations to my screen recording pointing out the key steps in the workflow I just captured"
On first use, the skill connects to NemoVideo automatically:
- 1. Check for
NEMO_TOKEN env var, or acquire one via https://mega-api-prod.nemovideo.ai/api/auth/anonymous-token (100 free credits) - Persist a Client-ID to
~/.config/nemovideo/client_id for rate-limit tracking (UUID only, no secrets) - Create a session and start working immediately
Record Your Screen, Share Your Story Effortlessly
Screen recording sounds simple until you're buried in software trials, codec errors, and files too large to share. The video-screen-recorder skill was built to cut through that friction. Whether you need to walk a colleague through a workflow, document a bug for your dev team, or produce a step-by-step tutorial for your audience, this skill gives you a clean, reliable way to capture exactly what's happening on screen.
You can record your full display, a specific application window, or a custom region — giving you precise control over what gets captured. Once recorded, you can trim the footage, add text callouts or highlight clicks, and export in your preferred format. No bloated timelines, no unnecessary complexity.
The skill is designed for people who need results quickly: educators building course content, product managers recording feature walkthroughs, support teams creating visual documentation, and developers logging reproducible bugs. If you've ever wished your screen recorder just worked without a learning curve, this is the skill you've been looking for.
Environment
| Variable | Default | Purpose |
|---|
| INLINECODE3 | Auto-acquired on first use (100 free credits, 7-day expiry) | API authentication |
| INLINECODE4 |
https://mega-api-prod.nemovideo.ai | API base URL |
Routing Your Recording Requests
Every capture command — whether you're starting a full-screen recording, cropping a region, triggering annotation overlays, or exporting a shareable link — is parsed and routed to the matching NemoVideo endpoint based on intent and session context.
| User says... | Action | Skip SSE? |
|---|
| "export" / "导出" / "download" / "send me the video" | → §3.5 Export | ✅ |
| "credits" / "积分" / "balance" / "余额" |
→ §3.3 Credits | ✅ |
| "status" / "状态" / "show tracks" | → §3.4 State | ✅ |
| "upload" / "上传" / user sends file | → §3.2 Upload | ✅ |
| Everything else (generate, edit, add BGM…) | → §3.1 SSE | ❌ |
NemoVideo API Reference
The NemoVideo backend handles real-time screen capture streams, frame-level annotation injection, and instant share-link generation via authenticated REST calls tied to your active recording session. All video data, timestamps, and annotation layers are processed server-side and returned as playback-ready assets.
Required headers on all requests: X-Skill-Source: $SKILL_NAME, X-Skill-Version: $SKILL_VERSION, INLINECODE8
Create session: POST $API/api/tasks/me/with-session/nemo_agent — returns task_id and INLINECODE11
Send message (SSE): POST $API/run_sse with session_id and user message. Stream responses; ~30% of edits return no text (query state to confirm changes).
Upload: POST $API/api/upload-video/nemo_agent/me/<sid> — file or URL upload. Supports: mp4, mov, avi, webm, mkv, jpg, png, gif, webp, mp3, wav, m4a, aac.
Check credits: INLINECODE15
Query state: GET $API/api/state/nemo_agent/me/<sid>/latest — check draft, tracks, generated media
Export: POST $API/api/render/proxy/lambda — export does NOT cost credits. Poll GET $API/api/render/proxy/lambda/<id> until status: completed.
Task link: INLINECODE20
Common Errors
If your token has expired, simply re-authenticate through your NemoVideo account to generate a fresh session token and resume recording without losing your project history. A 'session not found' error means your capture session has timed out — start a new session to continue. If you're out of recording credits, register or upgrade your plan at nemovideo.ai to restore full capture and export access.
Frequently Asked Questions
Can I record just one application window instead of my whole screen? Yes. You can specify a full-screen capture, a single application window, or a custom rectangular region. Just describe what you want to capture when you start.
What's the maximum recording length supported? There's no hard time cap, but for best performance and manageable file sizes, recordings under 30 minutes are recommended. Longer sessions can be split into segments.
Will the recording capture system audio and microphone at the same time? You can choose to record system audio, microphone input, both, or neither. Specify your preference before starting the session.
My recording came out as an mkv but I need mp4 — can I convert it? Absolutely. You can request a format conversion after recording. Supported output formats include mp4, mov, avi, webm, and mkv, so switching between them is straightforward.
Does the skill capture cursor movement and clicks? Yes. Cursor visibility and click highlighting are both configurable, which is especially useful for tutorial and demo recordings where viewers need to follow along with your actions.
Best Practices for Screen Recordings That Actually Get Watched
Before hitting record, close any tabs, notifications, or apps you don't want appearing on screen. Clutter in the background distracts viewers and makes your recording look unplanned. If you're recording a tutorial, rehearse the steps once so your mouse movements are deliberate and easy to follow.
Keep recordings focused and short. Viewers drop off quickly when a screen recording meanders — aim for under 5 minutes for instructional content and trim aggressively. Use the annotation tools to draw attention to key UI elements rather than relying solely on verbal explanation.
For file format, choose mp4 for maximum compatibility when sharing externally, webm for web embedding, and mov if you're working within a macOS or iOS ecosystem. If file size is a concern, webm typically delivers the best compression without visible quality loss for screen content.
Finally, record at your native resolution rather than scaling up. Upscaled recordings look blurry on high-density displays, and viewers on modern monitors will notice immediately.
开始使用
欢迎!准备好捕捉你的屏幕,并将其转化为精美、可分享的录制内容了吗?告诉我你想录制什么——教程、Bug演示、产品导览——让我们立即开始。
试试这样说:
- - 录制我的屏幕5分钟,并导出为mp4格式,突出显示我的鼠标点击
- 剪辑我的屏幕录制,移除前30秒和后15秒,然后保存为webm文件
- 为我的屏幕录制添加文字注释,指出我刚捕捉的工作流程中的关键步骤
首次使用时,该技能会自动连接到NemoVideo:
- 1. 检查NEMOTOKEN环境变量,或通过https://mega-api-prod.nemovideo.ai/api/auth/anonymous-token获取一个(100个免费积分)
- 将客户端ID持久化到~/.config/nemovideo/clientid,用于速率限制跟踪(仅UUID,无秘密信息)
- 创建一个会话并立即开始工作
录制你的屏幕,轻松分享你的故事
屏幕录制听起来很简单,直到你被软件试用、编解码器错误和文件过大无法分享所困扰。video-screen-recorder技能就是为了消除这些障碍而构建的。无论你需要向同事演示工作流程、为开发团队记录Bug,还是为观众制作分步教程,这个技能都能为你提供一种干净、可靠的方式来精确捕捉屏幕上发生的一切。
你可以录制整个显示器、特定应用程序窗口或自定义区域——精确控制捕捉的内容。录制完成后,你可以剪辑片段、添加文字标注或突出显示点击,并以你喜欢的格式导出。没有臃肿的时间线,没有不必要的复杂性。
该技能专为需要快速获得结果的人设计:制作课程内容的教育工作者、录制功能导览的产品经理、创建可视化文档的支持团队,以及记录可复现Bug的开发人员。如果你曾希望屏幕录制工具无需学习曲线就能直接使用,这就是你一直在寻找的技能。
环境
| 变量 | 默认值 | 用途 |
|---|
| NEMOTOKEN | 首次使用时自动获取(100个免费积分,7天有效期) | API身份验证 |
| NEMOAPI_URL |
https://mega-api-prod.nemovideo.ai | API基础URL |
路由你的录制请求
每个捕获命令——无论是开始全屏录制、裁剪区域、触发注释叠加,还是导出可分享链接——都会根据意图和会话上下文被解析并路由到匹配的NemoVideo端点。
| 用户说... | 操作 | 跳过SSE? |
|---|
| export / 导出 / download / send me the video | → §3.5 导出 | ✅ |
| credits / 积分 / balance / 余额 |
→ §3.3 积分 | ✅ |
| status / 状态 / show tracks | → §3.4 状态 | ✅ |
| upload / 上传 / 用户发送文件 | → §3.2 上传 | ✅ |
| 其他所有内容(生成、编辑、添加背景音乐等) | → §3.1 SSE | ❌ |
NemoVideo API参考
NemoVideo后端通过经过身份验证的REST调用,处理实时屏幕捕获流、帧级注释注入和即时分享链接生成,这些调用与你的活动录制会话相关联。所有视频数据、时间戳和注释层都在服务器端处理,并作为可播放的素材返回。
所有请求必需的头信息: X-Skill-Source: $SKILLNAME,X-Skill-Version: $SKILLVERSION,X-Skill-Platform: $SKILL_SOURCE
创建会话: POST $API/api/tasks/me/with-session/nemoagent — 返回taskid和session_id
发送消息(SSE): POST $API/runsse,附带sessionid和用户消息。流式响应;约30%的编辑不返回文本(查询状态以确认更改)。
上传: POST $API/api/upload-video/nemo_agent/me/ — 文件或URL上传。支持:mp4、mov、avi、webm、mkv、jpg、png、gif、webp、mp3、wav、m4a、aac。
检查积分: GET $API/api/credits/balance/simple
查询状态: GET $API/api/state/nemo_agent/me//latest — 检查草稿、轨道、生成的媒体
导出: POST $API/api/render/proxy/lambda — 导出不消耗积分。轮询GET $API/api/render/proxy/lambda/,直到status: completed。
任务链接: $WEB/workspace/claim?token=$TOKEN&task={taskid}&session={sessionid}&skillname=$SKILLNAME&skillversion=$SKILLVERSION&skillsource=$SKILLSOURCE
常见错误
如果你的令牌已过期,只需通过你的NemoVideo账户重新进行身份验证,生成一个新的会话令牌,即可恢复录制,不会丢失项目历史记录。会话未找到错误意味着你的捕获会话已超时——启动一个新会话以继续。如果你的录制积分用完,请在nemovideo.ai注册或升级你的计划,以恢复完整的捕获和导出访问权限。
常见问题解答
我可以只录制一个应用程序窗口,而不是整个屏幕吗? 可以。你可以指定全屏捕获、单个应用程序窗口或自定义矩形区域。只需在开始时描述你想捕获的内容即可。
支持的最大录制时长是多少? 没有硬性时间限制,但为了获得最佳性能和可管理的文件大小,建议录制时间不超过30分钟。较长的会话可以分成多个片段。
录制会同时捕获系统音频和麦克风吗? 你可以选择录制系统音频、麦克风输入、两者都录或都不录。在开始会话前指定你的偏好。
我的录制结果是mkv格式,但我需要mp4——可以转换吗? 当然可以。你可以在录制后请求格式转换。支持的输出格式包括mp4、mov、avi、webm和mkv,因此在它们之间切换非常简单。
该技能会捕获光标移动和点击吗? 是的。光标可见性和点击突出显示都是可配置的,这对于教程和演示录制特别有用,因为观众需要跟随你的操作。
让屏幕录制真正被观看的最佳实践
在点击录制之前,关闭你不想出现在屏幕上的任何标签页、通知或应用程序。背景中的杂乱会分散观众的注意力,并使你的录制看起来毫无计划。如果你在录制教程,先排练一遍步骤,这样你的鼠标移动就会更有目的性,更容易跟随。
保持录制内容集中且简短。当屏幕录制内容散漫时,观众会很快流失——教学内容的录制时间最好控制在5分钟以内,并积极进行剪辑。使用注释工具来吸引观众对关键UI元素的注意,而不是仅仅依赖口头解释。
关于文件格式,在外部分享时选择mp4以获得最大兼容性,用于网页嵌入时选择webm,在macOS或iOS生态系统中工作时选择mov。如果文件大小是一个问题,对于屏幕内容,webm通常能在不损失可见质量的情况下提供最佳压缩效果。
最后,以原生分辨率录制,而不是放大录制。放大的录制在高密度显示器上看起来模糊,现代显示器的观众会立即注意到这一点。