Let agents control a logged-in Chrome browser through Playwriter MCP
<p>Use Playwriter to give MCP-compatible agents a bounded, stateful Playwright control surface over an operator-approved Chrome session.</p>
npx skills add agentskillexchange/skills --skill let-agents-control-a-logged-in-chrome-browser-through-playwriter-mcp
Use Playwriter when an agent needs to inspect or operate a real browser session with existing logins, cookies, extensions, and tab state instead of launching a fresh automation browser. The workflow is to install the Chrome extension or start the bundled browser, create a Playwriter session, and run scoped Playwright snippets through the CLI or MCP interface while keeping sensitive actions under human review.
What this skill actually does
This is skill-shaped because the repeatable job is narrow: connect an agent to a supervised browser-control session, execute explicit browser steps, collect snapshots or page state, and stop at the operator’s approval boundary. It is not a generic browser automation framework listing, a scraping platform, or a replacement for Playwright; invoke it specifically when authenticated, stateful browser context is needed for an agent workflow.
Inputs and prerequisites: Playwriter Chrome extension and CLI; MCP-compatible client when using MCP mode; Chrome/Chromium browser session.
Setup notes: Install the Playwriter Chrome extension or start the bundled Playwriter browser, install the CLI with npm i -g playwriter, create a session with playwriter session new, and run scoped commands with playwriter -s -e ‘‘ or configure the MCP server for a compatible client.
Source and verification boundary: use https://playwriter.dev as the canonical reference before running the workflow; keep commands, API calls, CLI usage, and generated outputs reviewable against that upstream source.
Framework fit: publish this as a MCP workflow only when the operator can invoke the documented toolchain directly, rather than treating the upstream project as a generic product listing.