The Connector works with any client that supports local MCP servers — ones launched as a subprocess and spoken to over stdio. kuali mcp setup knows how to configure six of them out of the box.
$
kuali mcp setup --profile myschool --client codex
writes ~/.codex/config.toml — existing entries preserved
$
kuali mcp setup --profile myschool --client gemini-cli
writes ~/.gemini/settings.json
$
kuali mcp setup --profile myschool --client copilot-cli
writes ~/.copilot/mcp-config.json
$
kuali mcp setup --profile myschool --client vscode
writes .vscode/mcp.json in the current directory
This one is project-local rather than user-global, so you can scope the Connector to the repo you're working in. Open VS Code in that folder and Copilot Chat will pick it up.
For a client without a built-in helper that does support launching a local stdio MCP server, add an entry like this to the client's server list:
"mcpServers": {
"kuali": { "command": "/usr/local/bin/kuali",
"args": ["mcp", "--profile", "myschool"] } }
Setting KUALI_API_KEY in the server's environment overrides profile resolution entirely. Store keys with kuali setup instead — the Connector looks them up at start.
$
kuali mcp verify --client codex
$
kuali mcp verify --client gemini-cli
$
kuali mcp verify --client copilot-cli
$
kuali mcp verify --client vscode
ON THIS PAGE
<- All guidesChatGPT only supports hosted MCP servers — it can't launch a local binary. Use OpenAI Codex CLI to drive Kuali from an OpenAI model today.
Read the FAQ ->Ready-to-use prompts for curriculum, research, build apps, imports, workflow analysis, and reporting.
Browse prompts →Tips for Claude Desktop, Claude Code, Codex, Gemini, Copilot, and VS Code.
AI assistants →Skip the chat — every capability is also a plain kuali command, ready for scripts and CI.
Command reference →