Start with the two diagnostics, check the troubleshooting runbook, and know where to file bugs, ask questions, or report security issues.
$
kuali doctor # config, URL, auth, API reachability
$
kuali mcp verify # MCP client config, binary path, permissions
Pass --profile <name> to scope to a specific profile, or --client <name> to inspect an AI client's MCP configuration. The troubleshooting guide has a runbook for the usual failure modes.
Not yet — it's a transport mismatch, not a Kuali policy. ChatGPT only supports hosted MCP servers; the Connector is local, so your API keys and data stay on your computer. To drive Kuali from an OpenAI model right now:
$
kuali mcp setup --profile myschool --client codex
For questions about what data you should access, which apps you have permission for, or how your institution configures Kuali, your campus Kuali administrator is almost always the fastest path — ask your department contact or IT help desk if you're not sure who that is.
Open an issue in the kuali-connector repository and include:
$
kuali version
$
kuali doctor --profile <name> -o json
$
kuali mcp verify --client <claude-desktop|claude-code|…>
Plus your OS and version, the exact command or prompt, and what you expected versus what happened.
The Kuali Community forum (accessible via your institution's Kuali admin) is a good place for workflows, prompts, and automation recipes.
ON THIS PAGE
Email security@kuali.co for anything that might be a vulnerability. Please don't disclose publicly until we've had a chance to respond.
security@kuali.co
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 →
