Analysis Summary
The context-engine MCP server exposes a single tool that starts a context engine for a given project root directory. The primary risk is path traversal or arbitrary filesystem access if the projectRoot parameter is not validated, potentially allowing an attacker to point the engine at sensitive directories outside the intended project scope. Overall attack surface is narrow given the single tool.

you can add this to your README which would increase Credibility
Full Report: https://agentseal.org/mcp/context-engine