Agent Browser CLI
vercel-labs/agent-browser
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking…
Drives browser surfaces inside cmux from the command line: open sites, snapshot elements, act on refs, wait for state changes and extract data.
Install cmux Browser Automation
$ npx skills add disler/learning-cmux-with-agents --skill cmux-browser -a claude-codeProject install by default; add -g for ~/.claude/skills/.
$ gh skill install disler/learning-cmux-with-agents cmux-browser --agent claude-codeProject scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).
$ git clone --depth 1 https://github.com/disler/learning-cmux-with-agents.git skills-src && mkdir -p .claude/skills && cp -r skills-src/ai_docs/cmux-skills/cmux-browser .claude/skills/cmux-browser && rm -rf skills-srcUse ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.
Claude Code skills documentation · loads skills from .claude/skills/
Install the "cmux-browser" agent skill from https://github.com/disler/learning-cmux-with-agents/tree/main/ai_docs/cmux-skills/cmux-browser into .claude/skills/cmux-browser/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cmux-browser", then confirm the skill loads.Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$skill-installer install https://github.com/disler/learning-cmux-with-agents/tree/main/ai_docs/cmux-skills/cmux-browserType this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
$ npx skills add disler/learning-cmux-with-agents --skill cmux-browser -a codexProject install goes to .agents/skills/; add -g for ~/.codex/skills/.
$ gh skill install disler/learning-cmux-with-agents cmux-browser --agent codexProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/disler/learning-cmux-with-agents.git skills-src && mkdir -p .agents/skills && cp -r skills-src/ai_docs/cmux-skills/cmux-browser .agents/skills/cmux-browser && rm -rf skills-srcUse ~/.agents/skills/ instead of .agents/skills for a personal install.
Codex skills documentation · loads skills from .agents/skills/
Install the "cmux-browser" agent skill from https://github.com/disler/learning-cmux-with-agents/tree/main/ai_docs/cmux-skills/cmux-browser into .agents/skills/cmux-browser/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cmux-browser", then confirm the skill loads.Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add disler/learning-cmux-with-agents --skill cmux-browser -a cursorProject install goes to .agents/skills/; add -g for ~/.cursor/skills/.
$ gh skill install disler/learning-cmux-with-agents cmux-browser --agent cursorProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/disler/learning-cmux-with-agents.git skills-src && mkdir -p .cursor/skills && cp -r skills-src/ai_docs/cmux-skills/cmux-browser .cursor/skills/cmux-browser && rm -rf skills-srcUse ~/.cursor/skills/ instead of .cursor/skills for a personal install.
Cursor skills documentation · loads skills from .cursor/skills/, .agents/skills/, .claude/skills/, .codex/skills/
Install the "cmux-browser" agent skill from https://github.com/disler/learning-cmux-with-agents/tree/main/ai_docs/cmux-skills/cmux-browser into .cursor/skills/cmux-browser/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cmux-browser", then confirm the skill loads.Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gemini skills install https://github.com/disler/learning-cmux-with-agents.git --path ai_docs/cmux-skills/cmux-browser--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
$ npx skills add disler/learning-cmux-with-agents --skill cmux-browser -a gemini-cliProject install goes to .agents/skills/; add -g for ~/.gemini/skills/.
$ gh skill install disler/learning-cmux-with-agents cmux-browser --agent gemini-cliProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/disler/learning-cmux-with-agents.git skills-src && mkdir -p .gemini/skills && cp -r skills-src/ai_docs/cmux-skills/cmux-browser .gemini/skills/cmux-browser && rm -rf skills-srcUse ~/.gemini/skills/ instead of .gemini/skills for a personal install, then run /skills reload.
Gemini CLI skills documentation · loads skills from .gemini/skills/, .agents/skills/
Install the "cmux-browser" agent skill from https://github.com/disler/learning-cmux-with-agents/tree/main/ai_docs/cmux-skills/cmux-browser into .gemini/skills/cmux-browser/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cmux-browser", then confirm the skill loads.Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ gh skill install disler/learning-cmux-with-agents cmux-browserInstalls for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
$ npx skills add disler/learning-cmux-with-agents --skill cmux-browser -a github-copilotProject install goes to .agents/skills/; add -g for ~/.copilot/skills/.
$ git clone --depth 1 https://github.com/disler/learning-cmux-with-agents.git skills-src && mkdir -p .github/skills && cp -r skills-src/ai_docs/cmux-skills/cmux-browser .github/skills/cmux-browser && rm -rf skills-srcUse ~/.copilot/skills/ instead of .github/skills for a personal install. Commit .github/skills so cloud agent and code review can use it.
GitHub Copilot skills documentation · loads skills from .github/skills/, .claude/skills/, .agents/skills/
Install the "cmux-browser" agent skill from https://github.com/disler/learning-cmux-with-agents/tree/main/ai_docs/cmux-skills/cmux-browser into .github/skills/cmux-browser/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cmux-browser", then confirm the skill loads.GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
$ npx skills add disler/learning-cmux-with-agents --skill cmux-browser -a opencodeOpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
$ gh skill install disler/learning-cmux-with-agents cmux-browser --agent opencodeProject scope by default (.agents/skills/); add --scope user for a personal install.
$ git clone --depth 1 https://github.com/disler/learning-cmux-with-agents.git skills-src && mkdir -p .opencode/skills && cp -r skills-src/ai_docs/cmux-skills/cmux-browser .opencode/skills/cmux-browser && rm -rf skills-srcUse ~/.config/opencode/skills/ instead of .opencode/skills for a personal install.
OpenCode skills documentation · loads skills from .opencode/skills/, .claude/skills/, .agents/skills/
Install the "cmux-browser" agent skill from https://github.com/disler/learning-cmux-with-agents/tree/main/ai_docs/cmux-skills/cmux-browser into .opencode/skills/cmux-browser/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "cmux-browser", then confirm the skill loads.OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
- Skill name
cmux-browser- GitHub stars
- 115
- Token cost
- ~1.3k tokens
- SKILL.md length
- 316 words
- Files
- 11 (incl. references)
- Skills in repo
- 5
- Repo updated
- First seen
- Licence
- MIT
At a glance
Drives browser surfaces inside cmux from the command line: open sites, snapshot elements, act on refs, wait for state changes and extract data.
- Works in 6 steps: Open or target a browser surface. → Verify navigation with get url before… → Snapshot (--interactive) to get fresh… → …
- Automating a site inside a cmux browser pane instead of an external browser
- SKILL.md covers Core Workflow, Surface Targeting, Wait Support and Common Flows, plus 4 more sections
- Runs Shell scripts from its folder
What it does
The workflow is to open or target a browser surface, confirm navigation with get url, take an interactive snapshot to obtain fresh element refs, act with those refs (click, fill, type, select, press), wait for the page to change and snapshot again after any DOM or navigation change. Surfaces are addressed with short refs such as surface:N, found with cmux identify, and one surface should be kept per task. If get url returns an empty value or about:blank, the agent navigates first instead of waiting on load state.
Wait commands can target a selector, visible text, a URL fragment or a load state, each with a timeout. Example flows cover form submission, clearing an input and a stable loop of navigate, verify, wait, snapshot, act and snapshot. Reference notes cover commands, snapshot refs, authentication and saved state, proxy support, session management and video recording, and templates for authenticated sessions, capture workflows and form automation are provided.
When your agent uses it
- Automating a site inside a cmux browser pane instead of an external browser
- Filling and submitting a web form from the command line
- Waiting for a page element, text or URL change before extracting data
Example prompts
- “Open example.com/signup in a cmux browser surface and fill in the form.”
- “Wait until the dashboard URL loads, then take a snapshot and list the buttons.”
- “Save the authenticated state after I log in, so the next session skips the login.”
- “Extract the table on this page and record a video of the session.”
Requirements
- cmux with browser surfaces available
Workflow steps
6 steps, taken from the first numbered list in SKILL.md.
- Open or target a browser surface.
- Verify navigation with get url before waiting or snapshotting.
- Snapshot (--interactive) to get fresh element refs.
- Act with refs (click, fill, type, select, press).
- Wait for state changes.
- Re-snapshot after DOM/navigation changes.
What it can do on your machine
Read from SKILL.md and the folder at commit 6eaacab. It shows what the files ask for, not the result of running them.
Tool permissions
Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.
From allowed-tools in the SKILL.md frontmatter.
Runs code
Ships script files (Shell), which the agent can run.
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md.
From URLs in SKILL.md, links to its own repository left out.
Credentials
Names no API keys, tokens, secrets or passwords.
From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.
Context cost
cmux Browser Automation loads about 1.3k tokens when it runs, and up to ~4.9k if it reads all its reference files. Until then it costs about 44 tokens; SKILL.md has 316 words of instructions outside code blocks.
Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.
Safety
Auto-check passed
The automated check found no risky patterns in SKILL.md.
Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.
SKILL.md
The full file from disler/learning-cmux-with-agents at commit 6eaacab, republished under its MIT licence (© disler). 316 words, ~1,336 tokens.
Download SKILL.mdSave it as
.claude/skills/cmux-browser/SKILL.md (or your agent's skills folder). This skill also uses 10 other files; get the full folder from GitHub.- name
- cmux-browser
- description
- End-user browser automation with cmux. Use when you need to open sites, interact with pages, wait for state changes, and extract data from cmux browser surfaces.
Browser Automation with cmux
Use this skill for browser tasks inside cmux webviews.
Core Workflow
- Open or target a browser surface.
- Verify navigation with
get urlbefore waiting or snapshotting. - Snapshot (
--interactive) to get fresh element refs. - Act with refs (
click,fill,type,select,press). - Wait for state changes.
- Re-snapshot after DOM/navigation changes.
cmux --json browser open https://example.com
# use returned surface ref, for example: surface:7
cmux browser surface:7 get url
cmux browser surface:7 wait --load-state complete --timeout-ms 15000
cmux browser surface:7 snapshot --interactive
cmux browser surface:7 fill e1 "hello"
cmux --json browser surface:7 click e2 --snapshot-after
cmux browser surface:7 snapshot --interactiveSurface Targeting
# identify current context
cmux identify --json
# open routed to a specific topology target
cmux browser open https://example.com --workspace workspace:2 --window window:1 --jsonNotes:
- CLI output defaults to short refs (
surface:N,pane:N,workspace:N,window:N). - UUIDs are still accepted on input; only request UUID output when needed (
--id-format uuids|both). - Keep using one
surface:Nper task unless you intentionally switch.
Wait Support
cmux supports wait patterns similar to agent-browser:
cmux browser <surface> wait --selector "#ready" --timeout-ms 10000
cmux browser <surface> wait --text "Success" --timeout-ms 10000
cmux browser <surface> wait --url-contains "/dashboard" --timeout-ms 10000
cmux browser <surface> wait --load-state complete --timeout-ms 15000
cmux browser <surface> wait --function "document.readyState === 'complete'" --timeout-ms 10000Common Flows
Form Submit
cmux --json browser open https://example.com/signup
cmux browser surface:7 get url
cmux browser surface:7 wait --load-state complete --timeout-ms 15000
cmux browser surface:7 snapshot --interactive
cmux browser surface:7 fill e1 "Jane Doe"
cmux browser surface:7 fill e2 "jane@example.com"
cmux --json browser surface:7 click e3 --snapshot-after
cmux browser surface:7 wait --url-contains "/welcome" --timeout-ms 15000
cmux browser surface:7 snapshot --interactiveClear an Input
cmux browser surface:7 fill e11 "" --snapshot-after --json
cmux browser surface:7 get value e11 --jsonStable Agent Loop (Recommended)
# navigate -> verify -> wait -> snapshot -> action -> snapshot
cmux browser surface:7 get url
cmux browser surface:7 wait --load-state complete --timeout-ms 15000
cmux browser surface:7 snapshot --interactive
cmux --json browser surface:7 click e5 --snapshot-after
cmux browser surface:7 snapshot --interactiveIf get url is empty or about:blank, navigate first instead of waiting on load state.
Deep-Dive References
| Reference | When to Use |
|---|---|
| references/commands.md | Full browser command mapping and quick syntax |
| references/snapshot-refs.md | Ref lifecycle and stale-ref troubleshooting |
| references/authentication.md | Login/OAuth/2FA patterns and state save/load |
| references/authentication.md#saving-authentication-state | Save authenticated state right after login |
| references/session-management.md | Multi-surface isolation and state persistence patterns |
| references/video-recording.md | Current recording status and practical alternatives |
| references/proxy-support.md | Proxy behavior in WKWebView and workarounds |
Ready-to-Use Templates
| Template | Description |
|---|---|
| templates/form-automation.sh | Snapshot/ref form fill loop |
| templates/authenticated-session.sh | Login once, save/load state |
| templates/capture-workflow.sh | Navigate + capture snapshots/screenshots |
Limits (WKWebView)
These commands currently return not_supported because they rely on Chrome/CDP-only APIs not exposed by WKWebView:
- viewport emulation
- offline emulation
- trace/screencast recording
- network route interception/mocking
- low-level raw input injection
Use supported high-level commands (click, fill, press, scroll, wait, snapshot) instead.
Troubleshooting
js_error on snapshot --interactive or eval
Some complex pages can reject or break the JavaScript used for rich snapshots and ad-hoc evaluation.
Recovery steps:
cmux browser surface:7 get url
cmux browser surface:7 get text body
cmux browser surface:7 get html body- Use
get urlfirst so you know whether the page actually navigated. - Fall back to
get text bodyorget html bodywhensnapshot --interactiveorevalreturnsjs_error. - If the page is still failing, navigate to a simpler intermediate page, then retry the task from there.
© disler, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file
Files
SKILL.md and 10 other files (references) in ai_docs/cmux-skills/cmux-browser of disler/learning-cmux-with-agents.
- SKILL.md
- agents/openai.yaml
- references/authentication.md
- references/commands.md
- references/proxy-support.md
- references/session-management.md
- references/snapshot-refs.md
- references/video-recording.md
- templates/authenticated-session.sh
- templates/capture-workflow.sh
- templates/form-automation.sh
Open the folder on GitHubat commit 6eaacab
Compare with similar skills
cmux Browser Automation next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.
| Skill | Stars | Used in | Tokens | Auto-check | Licence | Repo updated |
|---|---|---|---|---|---|---|
| cmux Browser Automation this skilldisler/learning-cmux-with-agents | 115 | — | ~1.3k | Automated safety check: Pass | MIT | |
| Agent Browser CLIvercel-labs/agent-browser | 44k | 24 repos | ~864 | Automated safety check: Pass | Apache-2.0 | |
| Agent Browserquran/quran.com-frontend-next | 1.9k | 40 repos | ~3.3k | Automated safety check: Pass | None | |
| Web Access via Browser CDPeze-is/web-access | 9.1k | 4 repos | ~2.2k | Automated safety check: Pass | MIT | |
| Dev Browser AutomationMemTensor/MemOS | 12k | 3 repos | ~1.7k | Automated safety check: Pass | Apache-2.0 | |
| Electron App Automationvercel-labs/agent-browser | 44k | 5 repos | ~1.7k | Automated safety check: Pass | Apache-2.0 |
Similar skills
- Official
Agent Browser
quran/quran.com-frontend-next
Automates browser interactions for web testing, form filling, screenshots, and data extraction.
Web Access via Browser CDP
eze-is/web-access
Routes every web task, from searching to logged-in browsing, through a tiered choice of search, fetch, curl or a real Chrome or Edge session driven over CDP.
Dev Browser Automation
MemTensor/MemOS
Automates a real browser through short TypeScript scripts that keep page state between runs, for navigating, filling forms, taking screenshots and extracting data.
- Official
Electron App Automation
vercel-labs/agent-browser
Automates Electron desktop apps such as VS Code, Slack or Discord by connecting agent-browser to their Chrome DevTools Protocol port.
- Official
Next Dev Loop
vercel/next.js
Verify Next.js runtime behavior after editing app code. An agent skill from vercel/next.js.
More from disler/learning-cmux-with-agents
cmux Customization
disler/learning-cmux-with-agents
Customizes cmux for a user by editing cmux.json actions, workspace layouts, dock controls, tab bar buttons, settings and Ghostty terminal preferences.
cmux Contributor Dev Workflow
disler/learning-cmux-with-agents
Contributor rules for the cmux repo: tagged Debug builds through reload.sh, first-time setup, Xcode project normalization and per-tag sidebar extension points.
cmux Agent Surface Control
disler/learning-cmux-with-agents
Drives the cmux terminal multiplexer from natural language to open, inspect, prompt, read and close windows, workspaces, panes, surfaces and agent sessions.
cmux Testing Rules
disler/learning-cmux-with-agents
Testing rules for the cmux Swift codebase: Swift Testing as the default framework, a two-commit regression policy, and tests that check runtime behavior, not source text.
Related
Categories
Questions about cmux Browser Automation
What does cmux Browser Automation do?
Drives browser surfaces inside cmux from the command line: open sites, snapshot elements, act on refs, wait for state changes and extract data. The workflow is to open or target a browser surface, confirm navigation with get url, take an interactive snapshot to obtain fresh element refs, act with those refs (click, fill, type, select, press), wait for the page to change and snapshot again after any DOM or navigation change. Surfaces are addressed with short refs such as surface:N, found with cmux identify, and one surface should be kept per task.
When should I use cmux Browser Automation?
cmux Browser Automation fits situations like: automating a site inside a cmux browser pane instead of an external browser; filling and submitting a web form from the command line; waiting for a page element, text or URL change before extracting data.
How do I install cmux Browser Automation in Claude Code?
Run `npx skills add disler/learning-cmux-with-agents --skill cmux-browser -a claude-code`. Or copy the skill folder (ai_docs/cmux-skills/cmux-browser in disler/learning-cmux-with-agents) into .claude/skills/cmux-browser in your project. Claude Code loads it when a task matches its description.
How do I install cmux Browser Automation in Codex?
Run `npx skills add disler/learning-cmux-with-agents --skill cmux-browser -a codex`. Or copy the skill folder (ai_docs/cmux-skills/cmux-browser in disler/learning-cmux-with-agents) into .agents/skills/cmux-browser in your project. Codex loads it when a task matches its description.
Can I use cmux Browser Automation in Cursor, Gemini CLI or GitHub Copilot?
Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add disler/learning-cmux-with-agents --skill cmux-browser -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/cmux-browser, .gemini/skills/cmux-browser, .github/skills/cmux-browser and .opencode/skills/cmux-browser in your project.
What does cmux Browser Automation need to run?
Going by SKILL.md and its folder, cmux Browser Automation needs a shell for the scripts in its folder. Our summary lists: cmux with browser surfaces available.
Does cmux Browser Automation access the network?
SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.
Is cmux Browser Automation safe to install?
Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.
What licence does cmux Browser Automation use?
cmux Browser Automation is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does cmux Browser Automation use?
About 1.3k tokens (SKILL.md is roughly 5.3k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full. Its references folder adds about 3.5k tokens, read only when the agent opens those files.
What are the alternatives to cmux Browser Automation?
Skills that share tags, products or a category with cmux Browser Automation: Agent Browser CLI (vercel-labs/agent-browser, 44k stars), Agent Browser (quran/quran.com-frontend-next, 1.9k stars), Web Access via Browser CDP (eze-is/web-access, 9.1k stars) and Dev Browser Automation (MemTensor/MemOS, 12k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains cmux Browser Automation?
disler (a GitHub user) maintains it in disler/learning-cmux-with-agents, which has 115 GitHub stars. The repository holds 5 skills in this directory. The repository was last updated on June 29, 2026.
Source: disler/learning-cmux-with-agents on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.