agent-browser
agent-browser enables programmatic control of web browsers, letting you automate clicks, text input, page navigation, and other user interactions at scale. Built for integration with AI agents and automation frameworks, it simplifies complex browser workflows without manual intervention.
agent-browser automates form interactions by letting you programmatically click buttons, fill text fields, and submit forms on web pages. You can specify element selectors and input values via command-line arguments or API calls, enabling agent-browser to execute complex form workflows without manual intervention. This is especially useful for automating login sequences, data entry tasks, and multi-step web processes.
AI-generated summary based on this skill's SKILL.md
Install
fcakyon/claude-codex-settings/agent-browser · repository language: Python
git clone https://github.com/fcakyon/claude-codex-settings
cp -r claude-codex-settings/plugins/agent-browser/skills/agent-browser ~/.claude/skills/agent-browserFrequently asked questions
AI-generated answers based on this skill's SKILL.md and metadata
How do I click and fill forms on websites using agent-browser?
agent-browser automates form interactions by letting you programmatically click buttons, fill text fields, and submit forms on web pages. You can specify element selectors and input values via command-line arguments or API calls, enabling agent-browser to execute complex form workflows without manual intervention. This is especially useful for automating login sequences, data entry tasks, and multi-step web processes.
What is agent-browser and what can it do?
agent-browser is a browser automation tool that enables programmatic control of web browsers, letting you automate clicks, text input, page navigation, and other user interactions at scale. Built for integration with AI agents and automation frameworks, agent-browser simplifies complex browser workflows without manual intervention, supporting tasks like web scraping, form submission, and interactive testing.
Can agent-browser extract data from web pages programmatically?
Yes, agent-browser excels at extracting text and structured data from websites. It can parse page content, read accessibility trees, and capture snapshots of page state, allowing you to programmatically retrieve information from dynamic web pages. This makes agent-browser ideal for web scraping, data collection, and automated information gathering tasks.
How does agent-browser handle authentication and login workflows?
agent-browser supports authentication and form submission workflows by automating the steps needed to log into web applications. You can script login sequences—filling username and password fields, clicking submit buttons, and waiting for page transitions—enabling agent-browser to manage authenticated sessions for subsequent interactions on protected pages.
Does agent-browser support taking screenshots and capturing page state?
Yes, agent-browser can take screenshots of websites automatically and capture page state snapshots. This functionality lets you document page appearance at specific moments, verify visual changes, and create records of web interactions—useful for testing, monitoring, and debugging automated workflows.
Can agent-browser manage multiple browser sessions and tabs in parallel?
agent-browser supports managing multiple browser sessions and tabs in parallel, enabling you to run concurrent automation tasks across different pages or sites. This capability is valuable for large-scale web automation, distributed testing, and scenarios where you need to interact with multiple web contexts simultaneously.
SKILL.md
rendered from the published skill — quoted content, verbatim
agent-browser core
Fast browser automation CLI for AI agents. Chrome/Chromium via CDP, no Playwright or Puppeteer dependency. Accessibility-tree snapshots with compact @eN refs let agents interact with pages in ~200-400 tokens instead of parsing raw HTML.
Most normal web tasks (navigate, read, click, fill, extract, screenshot) are covered here. Load a specialized skill when the task falls outside browser web pages — see When to load another skill.
The core
(truncated - see the full file via the links below)
Read as markdown · JSON record · Browse the source repository
File tree — 13 files
plugins/agent-browser/skills/agent-browser/SKILL.md
plugins/agent-browser/skills/agent-browser/references/authentication.md
plugins/agent-browser/skills/agent-browser/references/commands.md
plugins/agent-browser/skills/agent-browser/references/profiling.md
plugins/agent-browser/skills/agent-browser/references/proxy-support.md
plugins/agent-browser/skills/agent-browser/references/session-management.md
plugins/agent-browser/skills/agent-browser/references/snapshot-refs.md
plugins/agent-browser/skills/agent-browser/references/trust-boundaries.md
plugins/agent-browser/skills/agent-browser/references/video-recording.md
plugins/agent-browser/skills/agent-browser/references/webgpu.md
plugins/agent-browser/skills/agent-browser/templates/authenticated-session.sh
plugins/agent-browser/skills/agent-browser/templates/capture-workflow.sh
plugins/agent-browser/skills/agent-browser/templates/form-automation.sh