Fetch web pages with Lightpanda and automate browser interactions with agent-browser.
npx skills add https://github.com/cagdotin/agents --skill browser-cagdotin
Please help me install this Agent Skill. Skill: browser Source: https://github.com/cagdotin/agents/tree/main/skills/tools/browser Command: npx skills add https://github.com/cagdotin/agents --skill browser-cagdotin
AI workflows often require both reading content from the web and performing interactive browser actions, which previously demanded switching tools or contexts. This Skill provides a single interface to fetch web pages quickly and automate browser interactions, enabling AI agents to access, interpret, and manipulate web content seamlessly.
Open a web page and fetch its readable content using Lightpanda, then optionally perform interactive actions with agent-browser.
High-intent search queries and answers about installing and using this skill.
To fetch web page content for AI agents, this Skill uses Lightpanda for fast read-only retrieval, providing clean output suitable for article reading and documentation extraction without requiring full browser overhead.
Yes, you can automate form interactions and UI testing by routing tasks to agent-browser, which handles clicking, filling forms, taking screenshots, and navigating multi-step flows beyond simple read-only fetching.
The best way to combine article reading with login flow testing is using unified routing guidance: fetch readable content with Lightpanda first, then switch to agent-browser to perform the interactive login and UI actions.
Yes, when a web page fails to fetch read-only content, the Skill provides fallback guidance to route the task to agent-browser, enabling full browser automation to handle pages requiring interaction or JavaScript rendering.
You need Lightpanda for fast read-only web fetching and agent-browser for interactive automation tasks like clicking and form filling; the Skill provides routing guidance within SKILL.md to direct tasks to the appropriate tool.
Using separate tools for reading versus interacting optimizes performance: Lightpanda retrieves readable content quickly without browser overhead, while agent-browser manages stateful sessions and complex UI automation for multi-step flows.
drvossInspect DOM state, network traffic, and Core Web Vitals in Chrome DevTools.
semicolon-devteamValidate test cases against issue acceptance criteria and generate verification reports.
runablehqAutomate browser control and web interaction in a real Chrome session.
lodekeeperScrape web content with tiered HTTP and headless browser automation.
