LLMs like Claude, Cursor, and GitHub Copilot are getting better at solving coding problems. But anyone who has ever written a business feature knows that the real challenge is not just writing the code. The code should follow your existing coding patterns, respect organization specific linting rules, use the right libraries, and fit into your architecture. So no matter how tempting it is for a developer to just copy paste the title and maybe description from the ticket in your favourite ticket management system its highly likely not going to produce code which is closer to the final version.

AWS has recently introduced the MCP server for AWS Serverless. You can read the official announcement here. I’ve been testing it using the latest MCP server with continue.dev plugin and Claude Sonnet 4 as the LLM backend. While the MCP server bundles a variety of tools, I want to focus on the three that stood out the most in my usage. 🔍 1. get_lambda_guidance-This tool helps determine if a given use case is suitable for AWS Lambda.