How do I install the Browser skill?

Run npx skillmds@latest add tomevault-io/browser in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Browser safe to use?

SkillMD's automated safety review verdict for this skill is WARNING. Independent scanners report: SkillSpector: PASS, Skill Scanner: WARNING. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Browser?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Browser free to use?

Yes. Installing skills from SkillMD is free, and the skill stays under its author's original license.

Who published Browser?

tomevault-io (@tomevault-io) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Browser by tomevault-io — Agent Skill

Name: Browser
Author: tomevault-io

File contents

Browser CLI Skill

Control a Chromium browser via CDP. Install: https://github.com/camhahu/browser

curl -fsSL https://raw.githubusercontent.com/camhahu/browser/main/packages/cli/install.sh | bash

Core Loop

browser open https://example.com  # Starts headless browser, shows outline

browser click l3                  # 1. Click by label from outline
browser text ".product-list"      # 3. Read content

browser stop                      # Always stop when finished

Prefer text over screenshots. Text is faster and more reliable. Use screenshots only when visual layout matters.

Commands

# Navigation
browser open <url>                # Open URL (auto-starts headless browser if needed)
browser navigate <url>            # Navigate current tab
browser back / forward / refresh  # History navigation
browser stop                      # Close browser

# Interaction
browser click <selector>          # Click by label, text, or CSS
browser type <text> [selector]    # Type into input (special keys: Enter, Escape, Tab)
browser hover <selector>          # Move mouse to element
browser drag <source> <target>    # Drag element to another element
browser select <selector> <value> # Select dropdown option
browser scroll <target>           # Scroll to top, bottom, or selector

# Content
browser outline                   # Show outline again (or use -a for full structure)
browser text [selector]           # Extract text content
browser screenshot [name]         # Capture screenshot (use when layout matters)

# Tabs
browser tabs                      # List all open tabs
browser use <tab-id>              # Switch to tab
browser close [tab-id]            # Close tab

# Debugging
browser console                   # Show console output
browser network                   # List network requests
browser eval <js>                 # Run JavaScript (shares scope between calls)

Navigation commands (open, click, navigate, back, forward, refresh) show an outline after completing.

For debugging or user intervention, start a headed browser with browser start --headed.

For software rendering, use browser start --headless --software-rendering. To pass through Chrome flags like Docker's --no-sandbox, use browser start --chrome-arg --no-sandbox.

Run browser --help for full command list.

Outline

Outline shows interactive elements with labels:

[l1] link "Products" [href=/products]
[l2] link "About" [href=/about]
[b3] button "Sign up"
[i4] input [type=email] [placeholder="Email"]

Selectors

browser click l3                 # Label from outline (fastest)
browser click "Sign up"          # Text match (exact first, then partial)
browser click ".btn-primary"     # CSS selector

CSS selectors: #id, .class, tag, [attr=value], parent > child

Persistent Profile

Logins, cookies, and browsing data are preserved across sessions in ~/.browser/profile/.

browser config clear-profile              # Logout, reset cookies
browser config set persistentProfile false # Disable persistence

Converted and distributed by TomeVault — claim your Tome and manage your conversions.

---name: browserdescription: Control a Chromium browser via CDP. Use when automating web browsing, scraping pages, testing web apps, monitoring network traffic, or interacting with page elements. Use when this capability is needed.---# Browser CLI SkillControl a Chromium browser via CDP. Install: https://github.com/camhahu/browser```bashcurl -fsSL https://raw.githubusercontent.com/camhahu/browser/main/packages/cli/install.sh | bash```## Core Loop```bashbrowser open https://example.com # Starts headless browser, shows outlinebrowser click l3 # 1. Click by label from outlinebrowser text ".product-list" # 3. Read contentbrowser stop # Always stop when finished```**Prefer `text` over screenshots.** Text is faster and more reliable. Use screenshots only when visual layout matters.## Commands```bash# Navigationbrowser open <url> # Open URL (auto-starts headless browser if needed)browser navigate <url> # Navigate current tabbrowser back / forward / refresh # History navigationbrowser stop # Close browser# Interactionbrowser click <selector> # Click by label, text, or CSSbrowser type <text> [selector] # Type into input (special keys: Enter, Escape, Tab)browser hover <selector> # Move mouse to elementbrowser drag <source> <target> # Drag element to another elementbrowser select <selector> <value> # Select dropdown optionbrowser scroll <target> # Scroll to top, bottom, or selector# Contentbrowser outline # Show outline again (or use -a for full structure)browser text [selector] # Extract text contentbrowser screenshot [name] # Capture screenshot (use when layout matters)# Tabsbrowser tabs # List all open tabsbrowser use <tab-id> # Switch to tabbrowser close [tab-id] # Close tab# Debuggingbrowser console # Show console outputbrowser network # List network requestsbrowser eval <js> # Run JavaScript (shares scope between calls)```Navigation commands (`open`, `click`, `navigate`, `back`, `forward`, `refresh`) show an outline after completing.For debugging or user intervention, start a headed browser with `browser start --headed`.For software rendering, use `browser start --headless --software-rendering`. To pass through Chrome flags like Docker's `--no-sandbox`, use `browser start --chrome-arg --no-sandbox`.Run `browser --help` for full command list.## OutlineOutline shows interactive elements with labels:```[l1] link "Products" [href=/products][l2] link "About" [href=/about][b3] button "Sign up"[i4] input [type=email] [placeholder="Email"]```## Selectors```bashbrowser click l3 # Label from outline (fastest)browser click "Sign up" # Text match (exact first, then partial)browser click ".btn-primary" # CSS selector```CSS selectors: `#id`, `.class`, `tag`, `[attr=value]`, `parent > child`## Persistent ProfileLogins, cookies, and browsing data are preserved across sessions in `~/.browser/profile/`.```bashbrowser config clear-profile # Logout, reset cookiesbrowser config set persistentProfile false # Disable persistence```---> Converted and distributed by [TomeVault](https://tomevault.io/claim/camhahu) — claim your Tome and manage your conversions.

Frequently asked questions

Browser