This site exposes an MCP server — a standard way for AI clients like Claude Desktop, Cursor, or Continue to plug into external tools and data. Connect once and your AI can chat with my resume, look up roles, and grade my fit against a job — without leaving the tool you already use.
https://schultz.ai/api/mcpJSON-RPC 2.0 over HTTPS POST. GET it in a browser for a self-describing response. No auth required (per-IP rate-limited).
Open claude_desktop_config.json (Settings → Developer → Edit Config) and merge in:
{
"mcpServers": {
"robert-schultz": {
"command": "npx",
"args": ["-y", "mcp-remote", "https://schultz.ai/api/mcp"]
}
}
}Restart Claude Desktop. You'll see a hammer icon when robert-schultz tools are available. Ask: "Grade Robert's fit against this JD: …".
Settings → MCP → Add new server. Or paste this directly into ~/.cursor/mcp.json:
{
"mcpServers": {
"robert-schultz": {
"url": "https://schultz.ai/api/mcp"
}
}
}Cursor supports remote HTTP MCP servers natively — no bridge needed.
Any client that speaks MCP works. For HTTP-native clients, point at the URL directly. For stdio-only clients (Continue, etc.), use the mcp-remote bridge as shown in the Claude Desktop snippet. Or just curl it:
curl -X POST https://schultz.ai/api/mcp \
-H 'content-type: application/json' \
-d '{"jsonrpc":"2.0","id":1,"method":"tools/list"}'get_resume
Full resume JSON — companies, roles, metrics, expertise, FAQ.
get_role
One role in detail. Pass a slug like 'meta' or 'amazon'.
ask_robert
Natural-language Q&A in his voice. Optional audience lens.
fit_assessment
Paste a JD; get a calibrated score + strengths + honest gaps.
A resume is a document. A resume agent is a conversation. By exposing my background through MCP, your AI can do the legwork — pull facts, grade fit, draft an intro — without you having to context-switch into another tab.
If you're a hiring manager: connect this and let your AI screen me first. If you're another engineer: this whole thing is on GitHub — fork it for your own resume.