chrome-debug-mcp
Related Servers
Alternatives to chrome-debug-mcp
- FlicenseBqualityBmaintenanceEnables LLMs to perform browser automation through the Playwright framework with Chrome DevTools Protocol support, connecting to existing Chrome instances for advanced web interactions and JavaScript execution.1252-
- AlicenseNot gradedqualityCmaintenanceAn MCP Server for Chrome DevTools, following the Chrome DevTools Protocol. Integrates with Claude Desktop and Claude Code.308MIT
Related Servers
- AlicenseNot gradedqualityDmaintenanceA Chrome DevTools Protocol-based MCP server that enables AI coding assistants to control browsers for JavaScript debugging, reverse engineering, web scraping, and API debugging.866 npm1Apache 2.0
- AlicenseAqualityAmaintenanceAn MCP server that connects AI agents to a running Chrome tab via the Chrome DevTools Protocol (CDP), enabling runtime debugging and page inspection.624153 npm1ISC
- FlicenseNot gradedqualityDmaintenanceA lightweight MCP server that enables AI assistants to control Chrome DevTools via CDP for debugging tasks like navigation, screenshots, and JavaScript execution.-
- AlicenseAqualityDmaintenanceA Model Context Protocol server that bridges MCP and the Debug Adapter Protocol, enabling AI coding agents to launch and control debug sessions, set breakpoints, step through code, and inspect variables.17MIT
- AlicenseNot gradedqualityDmaintenanceA fork of Chrome DevTools MCP that adds Streamable HTTP transport for multiple AI agents to concurrently control and inspect a live Chrome browser via the Model Context Protocol.16 npm1Apache 2.0
- AlicenseNot gradedqualityDmaintenanceMCP server that lets coding agents control and inspect live Chrome browsers, providing full Chrome DevTools capabilities for automation, debugging, and performance analysis.1,516,489 npmApache 2.0
TDQS
Scored across 35 tools
Most core browser and debugging tools are distinct (navigate, click, fill, evaluate), but there is notable overlap between lifecycle and state tools (stop_chrome vs close_instance, restart vs close, get_network_logs vs get_custom_events) and between stepping commands (step_over description incorrectly mentions 'step_over enters functions' while 'step_out' is mentioned but not defined). This ambiguity can cause agent misselection in edge cases.
Tool names mostly follow a consistent verb_noun pattern (open_tab, close_tab, list_tabs, set_breakpoint, resume, step_over). There are minor inconsistencies: 'webmcp_*' prefix is a different convention, and some verbs like 'restart', 'stop', 'close' create near-synonyms. Overall pattern is predictable and readable.
With 35 tools, the count is on the heavier side, but the server covers a broad scope: browser lifecycle, tab management, DOM interaction, debugging, performance, network, console, and experimental CDP. Most tools earn their place; a few like enable_proxy_auth and webmcp_* could be bundled, but none feel redundant.
The tool surface covers the full browser automation lifecycle (open/close instances and tabs, navigate, interact, inspect, debug, monitor performance, and network). Missing operations include no explicit 'step_out' despite being referenced, and limited screenshot options (no element-specific capture). Minor gaps that agents can work around with evaluate_js or send_cdp_command.