Hi! First of all, thank you for building this awesome MCP.
I tested it with various large language models (LLMs) and noticed a recurring issue.
LLMs don't know how to use the MCP properly
It would be extremely helpful to integrate a “skill” or recommended system prompt directly into the MCP.
Making this feature officially available would make the stealth-browser-mcp truly “plug-and-play” for all LLMs, thereby significantly reducing user frustration.
Thank you for considering this improvement and for your hard work on the project!
Hi! First of all, thank you for building this awesome MCP.
I tested it with various large language models (LLMs) and noticed a recurring issue.
LLMs don't know how to use the MCP properly
It would be extremely helpful to integrate a “skill” or recommended system prompt directly into the MCP.
Making this feature officially available would make the stealth-browser-mcp truly “plug-and-play” for all LLMs, thereby significantly reducing user frustration.
Thank you for considering this improvement and for your hard work on the project!