Cmux Capture logo

Cmux Capture

OrganizationPopular
manaflow-ai
cmux-capture

Screenshot or record a cmux window from the CLI, for a pull request, a bug report, or to see what the app actually looks like right now. Use when a change is visual, when a drag or an animation cannot be described in text, or when a reviewer needs evidence rather than a claim.

Overview

Publishermanaflow-ai
Repositorycmux
Skill namecmux-capture
Stars
27.5K
Forks
2.4K
Bundled files
1
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • 1 bundled files

    Scripts, templates, and references the model can read while it works. Files are read-only and never executed.

  • Open source

    Published by manaflow-ai on GitHub. Read the source before you install it.

Installation

Install the Cmux Capture AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/manaflow-ai/cmux/tree/main/skills/cmux-capture
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/manaflow-ai/cmux.git /tmp/cmux
mkdir -p .claude/skills
cp -r /tmp/cmux/skills/cmux-capture .claude/skills/cmux-capture
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Cmux Capture in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Cmux Capture on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Cmux Capture is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Screenshots and clips of cmux

cmux shot writes a png or a jpeg of a cmux window. cmux record writes an mp4 or a gif of one. Both capture only cmux's own windows, both go through the same permission-free path, and both work in a release build and inside CI. No Screen Recording permission is involved and none is ever requested.

bash
cmux shot --label settings-sheet                  # png of the frontmost window
cmux shot --jpg --max-width 1200 --out ~/before.jpg
cmux record start --gif --max-seconds 8 --label sidebar-drag
cmux record note "dragging the workspace"         # caption, drawn into the clip
cmux record stop

shot prints <width>x<height> <bytes> <path>. record prints <id> <state> <frames> <path>. Both take --json for the whole response.

Which one

Take a screenshot when the thing is still: a sheet, a menu, a layout, a label, a color. Record when the thing is a change over time: a drag, an animation, a reorder, two views settling in an order, a spinner that should stop. A screenshot of a drag shows nothing, and a clip of a static sheet is a large file a reviewer has to click to play.

Working from a clip to a still

--region x,y,w,h is the same rectangle in both commands, in window points from the window's top left. So a detail noticed in a clip can be shot with the same four numbers:

bash
cmux record start --region 0,0,420,900 --max-seconds 6   # watch the sidebar
cmux shot --region 0,0,420,900 --label sidebar-final     # keep the end state

Notes for an agent recording its own cmux

  • A recording stops itself at --max-seconds (default 15, maximum 120), so a clip is never left running by an agent that stops paying attention. Set it above the time the actions take or the clip ends early.
  • One at a time. A second record start is refused with conflict while one is running. cmux record list says whether anything is going, and cmux record status reports the frame count, with the effective frame rate in --json.
  • Caption as you go. cmux record note "<text>" draws text into the clip from that moment on. A clip has no step list beside it, so one note before each thing a reviewer should notice is worth more than a longer clip. --no-captions turns them off.
  • The clip is only cmux. Another app in front of the cmux window does not appear in the frames, so a capture does not have to own the screen.
  • Frames are timed by when they were captured, so a clip plays at the speed things happened even when the sampler fell behind the requested rate. Check fps_effective in --json if a clip looks slower than expected.
  • A gif is the shape to paste into a pull request, and --gif already picks the 8 fps and half scale that keep one small; an mp4 is sharper and needs a click to play. scripts/pr-media.py uploads either and converts an mp4 to a gif on the way.
  • Neither command replaces a directory or destroys an existing file before the new one is closed. --out on a path that is not a file is refused, and so is one whose extension does not match the format: --out clip.gif needs --gif.

In a CI dogfood tour

A dogfood scenario records with a record step wrapping the steps that matter, and captions with note, rather than by calling the CLI. The step table and the options are in cmux-testing/references/dogfood-scenarios.md.

Deep-dive references

ReferenceWhen to use
references/commands.mdEvery flag, the response shape, the error codes, window selection, limits

Bundled files

The model reads these on demand while the skill is loaded. They are exposed as readable files and are never executed.

Frequently asked questions

What does the Cmux Capture AI skill do?

Screenshot or record a cmux window from the CLI, for a pull request, a bug report, or to see what the app actually looks like right now. Use when a change is visual, when a drag or an animation cannot be described in text, or when a reviewer needs evidence rather than a claim.

Why use Cmux Capture on TypingMind?

Because you install it once and use it with any model. Cmux Capture is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Cmux Capture in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/manaflow-ai/cmux/tree/main/skills/cmux-capture. TypingMind reads its SKILL.md and bundles its files and installs it as a skill you can enable per chat.

Which AI models can use Cmux Capture?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Cmux Capture?

As many as you like. As long as a model supports skills, you can use Cmux Capture with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Cmux Capture AI skill free?

It is published on GitHub by manaflow-ai. Check the repository for licensing terms. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇