How to use your AI agent to manage your resumes with cloudresume.io

Most resume builders trap your work behind a UI you only visit when you're already job hunting. cloudresume.io is built differently: every action in the app is also exposed over the Model Context Protocol (MCP), so an AI agent like Claude or Cursor can do the work for you — from a chat window, an editor, or a script.

This post walks through connecting an agent to your account and the kinds of things it can do once it's wired up.

Why MCP

MCP is an open standard for letting language models call into external services through a small set of typed tools. Once an MCP server is configured, your agent can list, read, and modify data without you copy-pasting JSON or clicking through forms.

For cloudresume.io that means the same operations you'd run in the dashboard — creating a resume, editing a work entry, syncing GitHub, exporting a PDF, publishing a live link — are available to your agent as first-class tools.

Connect your agent

You'll need a cloudresume.io account. The MCP server lives at https://cloudresume.io/api/mcp and authenticates against your existing login.

Claude Code

In a terminal:

claude mcp add --transport http cloudresume https://cloudresume.io/api/mcp

The first call will prompt you to authorize the agent against your account. After that, Claude Code can call any of the cloudresume tools directly.

Claude.ai and Claude Desktop

For Claude Pro, Max, Team, and Enterprise plans you can add cloudresume.io as a custom connector straight from the app. Open Settings → Connectors and click Add custom connector:

Give it a name like cloudresume.io, paste https://cloudresume.io/api/mcp as the URL, and hit Add. The first call will prompt you to authorize the connector against your account. The same dialog is available in the Claude Desktop app on macOS and Windows, and once added the connector shows up in every conversation.

Cursor, VS Code, Windsurf

Add this to your mcp.json (or your editor's MCP settings):

{
  "cloudresume": {
    "type": "http",
    "url": "https://cloudresume.io/api/mcp"
  }
}

Restart the editor and the cloudresume tools will appear in your agent panel.

You can find the same setup snippets — already filled in with your personal server URL — under Settings → MCP server in the dashboard.

What your agent can do

Once connected, the agent has access to a set of tools that mirror the dashboard:

• list_resumes — see every resume on your account.
• get_resume — read the full JSON Resume document for one of them.
• create_resume / clone_resume — start a new resume from scratch or from an existing one.
• update_resume — patch any field, from a single bullet to a whole work entry.
• import_resume — pull in a JSON Resume file from elsewhere.
• sync_github — refresh GitHub repositories and contribution data on your profile.
• list_templates / set_template — switch the look without rewriting a word.
• generate_pdf — export a pixel-perfect PDF and get a download link back.
• publish_resume / unpublish_resume / list_publications / update_publication — manage your live cloudresume.io/r/yourname page.

Because all of these are exposed as plain tools, the agent can chain them together. You don't issue tool calls yourself — you describe the outcome you want.

Practical examples

Here are a few prompts that show what this looks like in practice.

Add a new role and ship the PDF

Add my new role as Senior Engineer at Acme Corp starting March 2026. Highlight that I'm leading the platform team. Then export an updated PDF.

The agent will:

1. Call list_resumes to find the right resume.
2. Call update_resume to insert the new work entry.
3. Call generate_pdf and hand you back the file.

Tailor a resume for a specific job

Clone my main resume into one called "Stripe SRE", then rewrite the summary and top three bullets to emphasize reliability and on-call experience.

The agent clones the resume with clone_resume, then patches the relevant fields with update_resume. You end up with a fresh, role-specific resume sitting next to your default one.

Keep GitHub in sync

Sync my GitHub repositories on my main resume and let me know which ones got added.

sync_github pulls in new public projects so the resume reflects what you've actually shipped lately. Useful to run before sending out applications.

Publish and share

Publish my main resume at the subdomain alex-johnson and give me the link.

The agent calls publish_resume, and you get back a live URL you can drop into a job application or your email signature.

Tips for getting good results

• Be specific about the resume. If you have more than one, mention it by title or role so the agent doesn't have to guess.
• Describe outcomes, not tool calls. "Add this role and refresh the PDF" works better than "call update_resume then generate_pdf".
• Let the agent read first. Asking it to read the current resume before editing usually produces cleaner diffs and avoids overwrites.
• Review changes in the dashboard. Every tool call writes to the same database the dashboard reads, so you can always open cloudresume.io to verify what changed.

Where this is going

The MCP server is the same surface the cloudresume.io dashboard uses internally. As we ship new features — cover letters, ATS scoring, job tailoring — they'll land as MCP tools too, so anything you can do in the app will be available to your agent on day one.

If you've been waiting for a resume builder that doesn't make you context-switch out of your editor, this is it. Sign up, connect your agent, and let it do the boring parts.