AI SummaryA Windsurf-native MCP server manager that enforces disciplined development workflows (plan → work → verify) while dynamically loading custom functions and external MCP servers on-the-fly. Ideal for developers building complex, multi-server AI agent systems who need strict code quality guardrails and runtime flexibility.
Install
Copy this and paste it into Claude Code, Cursor, or any AI assistant:
I want to add the "atlantis-mcp-server — Windsurf Rules" prompt rules to my project. Repository: https://github.com/ProjectAtlantis-dev/atlantis-mcp-server Please read the repo to find the rules/prompt file, then: 1. Download it to the correct location (.cursorrules, .windsurfrules, .github/prompts/, or project root — based on the file type) 2. If there's an existing rules file, merge the new rules in rather than overwriting 3. Confirm what was added
Description
MCP-compatible dynamic function manager; can be used with our Greenland terrain server or standalone
architecture overview
• this is a simple mcp (model context protocol) npx client and two server implementations: python and node • has the ability to add and remove things (namely dynamic functions and mcp servers) on the fly • ignore node-server folder for now, that is an mcp client • the python mcp sdk is located at /Users/reinman/work/mcp/python-sdk • the typescript node mcp sdk is located at /Users/reinman/work/mcp/typescript-sdk • server.py is the main python server that runs everything • NEVER TOUCH THE CLOUD MESSAGE PROTOCOL WITHOUT EXPLICIT APPROVAL BECAUSE IT CAN BREAK THE CLOUD SERVER
rules
• plan work first • do work • verify between steps • verify against plan • use meaningful logging so we can debug (quality over quantity) • whenever possible try to reuse existing code and logic • DO NOT BURY EXCEPTIONS • avoid over-engineering • avoid introducing new variables when existing ones will work fine • point out any problems you see w older code but do not correct unless related to task • when calling functions, carefully check order of parameters and data types
dynamic functions
• dynamic functions are stored in the dynamic_functions folder • see DynamicFunctionManager.py • dynamic functions can still be stored and retrieved even if invalid • dynamic functions invoked by tools/call • function_set must allow invalid Python
dynamic (mcp) servers
• adds the ability to host other mcp tools as if they are part of our own • these are simple JSON • see DynamicServerManager.py • server_set must allow invalid JSON • server_get_tools will get a list of mcp tools from a started mcp server (openweather and openweather.json is a good example)
Discussion
Health Signals
My Fox Den
Community Rating
Sign in to rate this booster
Works With
Any AI assistant that accepts custom rules or system prompts