mcpField
X
xfa

xfa

Remote
xfa.techSecurity0Updated 7w ago

About xfa

xfa is XFA’s remote MCP server for read-only queries about an organization’s devices, compliance, policies, software versions, and CVEs. It serves AI assistants with data scoped to the signed-in user’s organization and uses OAuth 2.0 authentication.

Key features

  1. 01Query organization and user details
  2. 02List devices and inspect individual devices
  3. 03Review compliance summaries and posture trends
  4. 04List organization policies
  5. 05Check software versions and related CVEs

Use cases

  1. 01Check device compliance across an organization
  2. 02Review changes in device posture over time
  3. 03Verify software versions and associated CVEs
  4. 04Query organization policies through an AI assistant

MCP support

The hosted remote MCP server exposes read-only tools for organizational devices, compliance, policies, software, and CVEs. Connections use OAuth 2.0 PKCE, with data scoped to the signed-in user’s organization.

Configuration

Remote servers need no local install. Pick your client, then copy the config or install in one click.

Endpointhttps://mcp.xfa.tech/mcp
Copy install command
Claude CodeTerminal
Cursor.cursor/mcp.json
VS Code.vscode/mcp.json
Claude Desktopclaude_desktop_config.json
Windsurf~/.codeium/windsurf/mcp_config.json

Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.

Tools

Details

From the project README

XFA MCP

XFA is a BYOD device-trust platform. This is XFA's remote Model Context Protocol server — query your organization's device posture, compliance, policies, and software/CVE status from Claude, ChatGPT, Cursor, and other AI assistants. Read-only.

This repository is the connector package for AI marketplaces. The MCP server itself is hosted by XFA at https://mcp.xfa.tech/mcp; nothing runs locally. Authentication is OAuth 2.0 (PKCE) — you sign in with your XFA account on connect.

Where it's published

SurfaceStatusWhere to submit / find
Official MCP Registry✅ Livetech.xfa/xfa · registry.modelcontextprotocol.io
Glama⏳ Propagating from registryglama.ai/mcp/servers
Smithery⏳ Propagating from registrysmithery.ai
PulseMCP⏳ Propagating from registrypulsemcp.com
mcp.so⏳ Propagating from registrymcp.so
Cursor Marketplace🕒 Submitted — pending approvalcursor.com/marketplace/publish
Claude Connectors Directory✅ Liveclaude.ai/directory/mcp-xfa-tech
ChatGPT app directory✅ Livechatgpt.com/plugins/…
Gemini / Antigravity CLI✅ Installable · ⏳ gallery auto-crawlgemini-cli-extension topic set; also via the MCP Registry
awesome-mcp-servers🕒 PR open — #12739Security section

Legend: ✅ live · ⏳ propagating (no action) · 🕒 pending. Update a row when its listing goes live.

Gemini note: the Gemini CLI gallery has no submission form — it crawls public repos tagged with the gemini-cli-extension GitHub topic daily (already set). Gemini CLI merged into Antigravity CLI (June 2026); Antigravity discovers MCP servers via the MCP Registry, where this server is already live, so no separate Antigravity submission is needed.

Install

Cursor

One-click:

cursor://anysphere.cursor-deeplink/mcp/install?name=xfa&config=eyJ1cmwiOiJodHRwczovL21jcC54ZmEudGVjaC9tY3AifQ==

Or add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "xfa": { "url": "https://mcp.xfa.tech/mcp" }
  }
}
Claude

Settings → Connectors → Add custom connector → URL https://mcp.xfa.tech/mcp.

ChatGPT

Settings → Connectors → Add → MCP server URL https://mcp.xfa.tech/mcp.

Gemini CLI

Install the extension:

gemini extensions install https://github.com/gl-ventures/xfa-mcp

The bundled gemini-extension.json points at the remote server; Gemini discovers OAuth from the server metadata and prompts you to sign in on first use.

VS Code / Windsurf / Zed / other MCP clients

Point the client at the remote URL https://mcp.xfa.tech/mcp (streamable HTTP / SSE, OAuth 2.0).

Tools

All tools are read-only. The live server is the source of truth — clients fetch the current tool list from the endpoint on every connection, so this list may lag the deployed server. See the Connect an AI assistant docs for the maintained reference.

Your organization (scoped to the signed-in user's org)

ToolDescription
get_organizationGet your XFA organization
get_current_userGet the signed-in user
list_devicesList devices (active in the last 30 days)
get_deviceGet a single device
get_compliance_summaryOrg-wide compliance summary
get_posture_trendsPosture trends over time
list_policiesList policies

Software & vulnerabilities (XFA's tracked-software catalog)

ToolDescription
get_latest_versionLatest known version of a piece of software
list_versioned_software_catalogList the software XFA tracks
get_software_version_infoStatus of a specific software version
get_cves_for_versionCVEs affecting a software version

Authentication

OAuth 2.0 with PKCE (S256), scope mcp:read. On connect you are redirected to XFA to authorize; no API keys or tokens are stored in this package. The client auto-discovers the OAuth endpoints from the server's already-published metadata:

  • https://mcp.xfa.tech/.well-known/oauth-protected-resource
  • https://mcp.xfa.tech/.well-known/oauth-authorization-server

Support

Maintainers

The MCP Registry workflow validates server.json on every change and publishes to the official MCP Registry on pushes to main (or via Run workflow). It authenticates by DNS against the v=MCPv1 TXT record on the xfa.tech apex.

To release a new version: bump version in server.json, merge to main.

Required repo secret: MCP_REGISTRY_KEY_PEM — the Ed25519 private key PEM (pairs with the DNS TXT record). Keep the matching key backed up in a password manager.

License

MIT — see LICENSE.

Comments

Markdown supported

New to MCP?

From what the protocol is to installing your first server, these answers cover most first questions.

An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.