After installing chrome-devtools-cli, is it unnecessary to use chrome devtools mcp?

[duguyue8]

I normally use an agent to operate the Chrome browser to perform some automation tasks. I need to reuse the websites I'm already logged into without having to log in again.

[swimmwatch]

chrome-devtools-cli and chrome-devtools-mcp overlap, but they are aimed at different workflows.

If your main need is to have an agent reuse websites where you're already logged in, Chrome DevTools MCP can attach to a running Chrome instance with remote debugging enabled (--browser-url=...). In that model, the agent is inspecting/controlling that browser session, so it can use the same logged-in profile you intentionally launched for that purpose.

Playwright MCP is a different workflow: it exposes Playwright-style browser automation tools and usually manages browser contexts more like test automation. It can persist state with a user data directory, but it is not the same as attaching DevTools to your everyday Chrome session.

So installing the CLI does not make MCP unnecessary; use the CLI when you want terminal-driven commands, and use MCP when you want your coding agent to call browser tools directly.

For completeness, if someone specifically wants Playwright MCP tool semantics backed by CloakBrowser Chromium, CloakBrowser MCP is a bridge for that use case. It is still a Playwright MCP-style workflow, not a replacement for Chrome DevTools MCP attach/reuse. Disclosure: I maintain CloakBrowser MCP.