Guide · Agent skill
How to use the ego-browser agent skill for web automation
Updated · written with AI from the project's own documentation and checked against our catalogue data
The ego-browser skill tells an AI agent to do browser work in ego-browser (ego-lite), a Chromium browser built for both people and AI agents. This guide explains what a skill is, how to install this one, how to ask for browser tasks in plain language, and what to watch for before letting an agent act inside your logged-in sessions.
ego-browser Web Automationcitrolabs/ego-lite · MIT Download the skill Quick start
- 1Install the ego-browser program. Install ego-browser (ego-lite) and confirm it is on your PATH. The skill cannot work without it.
- 2Download the skill ZIP. Get the ZIP from the skill page on this site and unzip it.
- 3Place the skill folder. Put it in ~/.claude/skills/ for all projects, or in .claude/skills/ inside one project. Alternatively, upload it under Settings, Capabilities, Skills in the Claude apps.
- 4Confirm the guide runs. Run ego-browser skill and check that the output ends with the end-of-guide line.
- 5Ask for a browser task. Describe the site, the action and the result you want in plain language.
- 6Review the result. Check screenshots, extracted data and any form input before accepting the work, especially on signed-in accounts.
What an agent skill is and what this one does
An agent skill is a folder of instructions and optional scripts that an AI agent loads when a task matches it. The SKILL.md file in this folder is the entry point the agent reads.
With this skill, the agent is told to use ego-browser for browser tasks by default. That covers opening and operating websites, filling forms, clicking buttons, taking screenshots, extracting page data, signing in, and web app testing such as QA and bug investigation.
- Open and operate websites
- Fill forms and click buttons
- Take screenshots and extract page data
- Sign in and use your logged-in sessions
- Test web apps, reproduce bugs and review app quality
Install the ego-browser skill
Download the ZIP from the skill page on this site, then unzip it. Put the resulting skill folder in one of these places: ~/.claude/skills/ for your user account, or .claude/skills/ inside a specific project.
You can also upload the skill in the Claude apps. Go to Settings, then Capabilities, then Skills, and add it there.
The skill needs the ego-browser program (ego-lite) installed and available on your PATH. If the agent reports that ego-browser is not found, the skill's install reference covers installing or upgrading it.
Check that the guide loads
Before any browser task, the skill has the agent run ego-browser's built-in usage guide. You can run the same command yourself to confirm the program responds.
The guide ends with a line starting [ego-browser:skill] end of guide. If that line is missing, the output was cut off. Save it to a file and read the file in parts.
- Run it again after ego-browser upgrade, or when any output tells you to.
ego-browser skillHow to trigger the skill with natural requests
You do not need special syntax. Describe the browser task you want done. The skill's description covers browser automation, web app testing, QA and app-quality review, so requests like these should match it.
Name the site and the goal. If the task involves an account you are signed into, say so, because the agent can use your logged-in sessions.
- Open the pricing page on our site, take a screenshot and tell me what changed since last week.
- Fill out the contact form on this page with my details from the notes file and stop before submitting.
- Go to my repo's issues page and list the open issues with labels.
- Test the signup flow in our staging app and report any errors you find.
Helper scripts included with the skill
The skill folder includes small scripts for specific sites. They add shortcuts for common lookups instead of having the agent work out the page each time.
GitHub: repo stats, searching repositories, and listing open issues. Google: search results extraction and autocomplete suggestions. X (Twitter): extracting post text. Each site folder also has notes describing the page structure it expects.
- learnings/github: repo-stats.js, search-repos.js, open-issues.js
- learnings/google: search-extract.js, autocomplete.js
- learnings/x-com: extract-post.js
Safety, scripts and network use
This skill runs scripts and uses the network. It visits the sites you ask about, and it can act inside sessions you are already logged into. Only use it where you are comfortable with that.
Be careful with clearing cookies, cache or storage. Some clears affect the whole browser profile and can sign you out or delete data. The skill's clearing-state reference should be read before any clear.
The helper scripts read page elements that may change without notice, so results can break after a site redesign. Check that the target site's terms allow automated access.
- Review what the agent plans to post, buy or submit before it does so.
- Avoid asking for clears of cookies or storage unless you have read the clearing reference.
- Do not use it on sites whose terms prohibit automated access.
What people use it for
- Checking a page for changes and taking screenshots over time
- Filling out repetitive web forms while you review before submit
- Listing open GitHub issues or searching repositories without writing scraping code
- Pulling Google search results or autocomplete suggestions for research
- Reproducing a bug in a web app and describing the steps taken
Frequently asked questions
How do I install the ego-browser skill?
Download the ZIP from the skill page, unzip it, and place the folder in ~/.claude/skills/ or a project's .claude/skills/ folder. You can also upload it in the Claude apps under Settings, Capabilities, Skills.
What do I need before the ego-browser skill works?
You need the ego-browser program (ego-lite) installed and on your PATH, network access to the sites you visit, and logged-in browser sessions for any personal accounts the agent should use.
Is the ego-browser skill free?
The source for this item does not state pricing or licence terms beyond the MIT licence listed for the repository. Check the repository's licence file for the exact terms.
What is the difference between ego-browser and a normal browser?
According to the skill description, ego-browser is a Chromium browser designed for both human users and AI agents, and agents can use your logged-in websites and context through it. The skill tells the agent to prefer it over built-in browsers or other web tools.
Why does the agent say ego-browser is not found?
The program is probably not installed or not on your PATH. The skill's install reference covers installing or upgrading ego-browser, and you can confirm it with the ego-browser skill command.
Is it safe to let an agent use my logged-in accounts?
It can read and act within sessions you are signed into, which can lead to unintended posts, purchases or data exposure. Review planned actions, avoid unnecessary cookie or storage clears, and only use it where you accept that risk.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
Use this skill if you want an agent to handle repetitive browser tasks in a browser you already use, or to test and review web apps, and you are comfortable with it acting in your logged-in sessions. Pick a different tool if you do not want an agent acting inside your accounts, or if the sites you need prohibit automated access.
Download the skill