I need a small MCP server written in Python that can speak fluently with the Xiaozhi MCP client (spec here – https://ccnphfhqs21z.feishu.cn/wiki/HiPEwZ37XiitnwktX13cEM5KnSb) and forward each incoming AI-generated message straight into a relevant designated Slack channel. FastMCP (https://github.com/jlowin/fastmcp) already does most of the protocol work, so feel free to fork or adapt it rather than starting from scratch. The finished server must run on my pythonanywhere.com account; I will supply credentials once the code is ready for deployment. The Slack side is simple: I already have a Slack app with the required OAuth scopes and a bot token, so you only need to wire your server to its Web API endpoint. No receiving or slash-command handling is needed—outbound “send-only” traffic is enough. The AI bot itself decides the message text; your server just passes it through unchanged. Deliverables • Clean Python code (3.9+), preferably using FastAPI or Flask alongside the FastMCP core • Working deployment on pythonanywhere.com, reachable over HTTPS • A short README covering setup, environment variables, and how to add new Slack channels • Quick confirmation test: when I trigger the Xiaozhi client, the message must appear in Slack within a second or two That’s all—once those boxes are ticked, we’re done.