agent-browser
This skill equips your AI agents with the ability to control web browsers programmatically, handling clicks, form submissions, and page navigation as part of automated workflows. Built on a Go-based engine that communicates via WebSocket, it integrates seamlessly into multi-turn agent conversations and supports custom tool extensions. Perfect for automating repetitive web tasks or building intelligent browser automation into your agent applications.
agent-browser is a skill that equips your AI agents with the ability to control web browsers programmatically, handling clicks, form submissions, and page navigation as part of automated workflows. It enables you to automate clicking buttons and filling forms by providing programmatic control over browser interactions, allowing agents to interact with web elements just as a human would. The skill is built on a Go-based engine that communicates via WebSocket, making it easy to integrate into multi-turn agent conversations.
AI-generated summary based on this skill's SKILL.md
Decision gist · record as of 2026-07-23
agent-browser is a skill that equips your AI agents with the ability to control web browsers programmatically, handling clicks, form submissions, and page navigation as part of automated workflows. It enables you to automate clicking buttons and filling forms by providing programmatic control over browser interactions, allowing agents to interact with web elements just as a human would. The skill is built on a Go-based engine that communicates via WebSocket, making it easy to integrate into multi-turn agent conversations.
Use it when
- Yes, agent-browser supports extracting data and taking screenshots from web pages through a command-line interface.
- agent-browser manages authentication and persistent sessions for automated workflows, enabling you to log in to websites automatically.
Verify before relying
Read SKILL.md below before installing (11 files). Open directory: indexed for reading, not audited.
Install
harnessclaw/harnessclaw-engine/agent-browser · repository language: Go
Open directory. Skills are indexed for reading, not audited. Review a skill's body before installing it.
Frequently asked questions
AI-generated answers based on this skill's SKILL.md and metadata
What is agent-browser and how does it automate clicking buttons and filling forms?
agent-browser is a skill that equips your AI agents with the ability to control web browsers programmatically, handling clicks, form submissions, and page navigation as part of automated workflows. It enables you to automate clicking buttons and filling forms by providing programmatic control over browser interactions, allowing agents to interact with web elements just as a human would. The skill is built on a Go-based engine that communicates via WebSocket, making it easy to integrate into multi-turn agent conversations.
Can agent-browser extract data and take screenshots from web pages via command-line interface?
Yes, agent-browser supports extracting data and taking screenshots from web pages through a command-line interface. This capability allows you to programmatically capture visual representations of websites and pull structured data from web pages as part of your automated workflows. The CLI-based approach makes it straightforward to integrate screenshot capture and data extraction into scripts and agent applications without requiring a graphical interface.
How does agent-browser handle login to websites automatically for browser session management?
agent-browser manages authentication and persistent sessions for automated workflows, enabling you to log in to websites automatically. The skill maintains browser session state across interactions, allowing your agents to authenticate once and then perform subsequent actions within that authenticated session. This persistent session management is essential for automating workflows that require users to be logged in, such as accessing account-specific data or performing authenticated transactions.
Is agent-browser suitable for headless browser testing and test web apps with CLI?
Yes, agent-browser is well-suited for headless browser testing and testing web applications via command-line interface. The skill enables you to test web applications and verify functionality across different viewports and devices without requiring a graphical display. Its CLI integration makes it ideal for automated testing pipelines, continuous integration workflows, and programmatic verification of web app behavior across various configurations.
What makes agent-browser a browser automation tool without selenium?
agent-browser provides chromium automation capabilities as an alternative to Selenium-based approaches. Rather than relying on Selenium's architecture, agent-browser is built on a Go-based engine that communicates via WebSocket, offering a more lightweight and agent-friendly approach to browser automation. This design makes it particularly well-suited for integrating intelligent browser automation into your agent applications while supporting custom tool extensions for specialized use cases.
Can agent-browser be used for mobile browser testing and dark mode testing?
agent-browser supports testing web applications across different viewports and devices, which includes mobile browser testing scenarios. This capability allows you to verify how your web applications behave on various screen sizes and device configurations. While the skill's primary focus is on automating browser interactions and testing functionality, its viewport flexibility enables you to test responsive design and device-specific behaviors as part of your comprehensive testing strategy.
SKILL.md
Rendered from the published skill. Quoted content, verbatim.
Browser Automation with agent-browser
The CLI uses Chrome/Chromium via CDP directly. Install via npm i -g agent-browser, brew install agent-browser, or cargo install agent-browser. Run agent-browser install to download Chrome. Run agent-browser upgrade to update to the latest version.
Core Workflow
Every browser automation follows this pattern:
- Navigate:
agent-browser open <url> - Snapshot:
agent-browser snapshot -i(get element refs like@e1,@e2) - Interact: Use refs to click, fill, select
- Re-snapshot: After navigation or DOM changes, get fresh
(truncated - see the full file via the links below)
File tree — 11 files
internal/engine/agent/builtin/browser_agent/resources/agent-browser/SKILL.md
internal/engine/agent/builtin/browser_agent/resources/agent-browser/references/authentication.md
internal/engine/agent/builtin/browser_agent/resources/agent-browser/references/commands.md
internal/engine/agent/builtin/browser_agent/resources/agent-browser/references/profiling.md
internal/engine/agent/builtin/browser_agent/resources/agent-browser/references/proxy-support.md
internal/engine/agent/builtin/browser_agent/resources/agent-browser/references/session-management.md
internal/engine/agent/builtin/browser_agent/resources/agent-browser/references/snapshot-refs.md
internal/engine/agent/builtin/browser_agent/resources/agent-browser/references/video-recording.md
internal/engine/agent/builtin/browser_agent/resources/agent-browser/templates/authenticated-session.sh
internal/engine/agent/builtin/browser_agent/resources/agent-browser/templates/capture-workflow.sh
internal/engine/agent/builtin/browser_agent/resources/agent-browser/templates/form-automation.sh
Let your AI agent find skills like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 56,283 agent skills by what they can do, searchable in plain language.
wish › “Automate browser interactions like clicking, filling forms, and navigating websites programmatically”
Give your agent the search over MCP, or paste the wish link into any chat. No install? Search from any chat →
Related skills
Agent Browser equips your AI agent with the ability to explore and interact with web content in real time. Navigate between pages, retrieve information, and perform actions on websites as if a human were at the controls. Perfect for agents that need autonomous web research and data gathering.
agent-browser enables you to script and automate browser-based workflows directly from the command line. Execute web interactions, navigate pages, and handle complex tasks without manual intervention. Built for developers who need reliable programmatic control over web automation.
Agent Browser equips Claude with the ability to visit web pages, extract content, and perform interactions like clicking and form submission—all through code. This skill transforms Claude into an autonomous web agent capable of gathering information and completing tasks across the internet without manual intervention.
agent-browser enables Claude Code to interact with web pages programmatically, handling clicks, form submissions, and navigation tasks. This skill bridges AI agents and browser automation, letting you build workflows that manipulate web content dynamically. Perfect for testing, data collection, and repetitive web tasks.
Agent Browser equips your AI agents with the ability to explore web content and perform actions on live websites in real time. Perfect for automating research, data collection, and web-based tasks that require intelligent navigation and interaction.
agent-browser lets you orchestrate automated browser interactions through a remote HTTP interface, enabling your applications to navigate websites, populate forms, trigger actions, and harvest information at scale. Built for sandbox deployment, it provides streaming event handling and session management across multiple coding agent platforms.
More skills Agent Browser (NOASSERTION) · browser (MIT) · Agent Browser (unlicensed) · core (Apache-2.0) · agent-browser (Apache-2.0) · Agent Browser (unlicensed) · agent-browser (MIT) · Agent Browser (AGPL-3.0)