agent-browser
agent-browser enables agents to interact with web pages autonomously, handling form submissions, element clicks, and multi-step navigation workflows. Built for cross-repository engineering practices, it provides the scaffolding needed to turn repetitive browser tasks into reliable automated processes.
agent-browser enables agents to interact with web pages autonomously, handling form submissions, element clicks, and multi-step navigation workflows. Built for cross-repository engineering practices, it provides the scaffolding needed to turn repetitive browser tasks into reliable automated processes. The tool is distributed under the MIT license, making it freely available for both open-source and commercial projects.
AI-generated summary based on this skill's SKILL.md
Install
onsager-ai/dev-skills/agent-browser · repository language: Shell
git clone https://github.com/onsager-ai/dev-skills
cp -r dev-skills/skills/agent-browser ~/.claude/skills/agent-browserFrequently asked questions
AI-generated answers based on this skill's SKILL.md and metadata
What is agent-browser used for?
agent-browser enables agents to interact with web pages autonomously, handling form submissions, element clicks, and multi-step navigation workflows. Built for cross-repository engineering practices, it provides the scaffolding needed to turn repetitive browser tasks into reliable automated processes. The tool is distributed under the MIT license, making it freely available for both open-source and commercial projects.
Can agent-browser automate website interactions with a CLI?
Yes. agent-browser is designed to automate website interactions with CLI support, allowing you to programmatically fill forms, click buttons, and navigate pages without manual intervention. This makes it ideal for scripting repetitive web tasks and integrating browser automation into larger workflows or agent-based systems.
How does agent-browser handle login to websites and extract data?
agent-browser manages authenticated sessions and persists login state across multiple browser sessions, enabling you to maintain user credentials and session data throughout your automation workflow. Once authenticated, agent-browser can extract data and content from websites programmatically, allowing you to retrieve information from protected pages and authenticated areas of web applications.
What testing capabilities does agent-browser provide?
agent-browser supports testing web applications by simulating user actions and verifying results, making it suitable for automated quality assurance workflows. You can script complex user interactions—such as form fills, button clicks, and multi-step navigation—then validate that the application responds as expected, all without manual browser operation.
Can agent-browser capture visual evidence of web pages?
Yes. agent-browser can capture visual evidence through screenshots and recordings for debugging or documentation purposes. This feature is particularly useful when you need to document test results, create visual records of web scraping operations, or troubleshoot issues by reviewing snapshots of page state at different points in your automation workflow.
What makes agent-browser suitable for programmatic web interaction?
agent-browser is a programmatic web interaction tool that combines form filling, element clicking, navigation, and data extraction into a unified interface. Its design supports both simple single-page tasks and complex multi-step workflows, making it adaptable for scenarios ranging from web scraping and form automation to comprehensive web application testing and agent-driven browser control.
SKILL.md
rendered from the published skill — quoted content, verbatim
Browser Automation with agent-browser
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 refs
```bash agent-browser open https://example.com/form agent-browser snapshot -i
Output: @e1 [input type="email"], @e2 [input type="password"], @e3 [button] "Submit"
agent-browser fill @e1 "user@example.com" agent-browser fill @e2 "password123" agent-browser click
(truncated - see the full file via the links below)
Read as markdown · JSON record · Browse the source repository
File tree — 10 files
skills/agent-browser/SKILL.md
skills/agent-browser/references/authentication.md
skills/agent-browser/references/commands.md
skills/agent-browser/references/proxy-support.md
skills/agent-browser/references/session-management.md
skills/agent-browser/references/snapshot-refs.md
skills/agent-browser/references/video-recording.md
skills/agent-browser/templates/authenticated-session.sh
skills/agent-browser/templates/capture-workflow.sh
skills/agent-browser/templates/form-automation.sh