MCP Server

Use Verity directly inside any MCP-compatible AI client — Claude, Codex, Cursor, and more.

The Verity MCP server lets any AI client query coverage policies, look up codes, and check prior auth requirements in real time — without writing any code.

MCP vs Agent Tools API — The MCP server is for AI assistants used interactively by developers and analysts. The Agent Tools API is for production agents embedded in your application, with audit logging, PHI protection, and human review routing built in.

Installation

Clone and build the server:

git clone https://github.com/backworkai/verity_mcp.git
cd verity_mcp
npm install
npm run build

Get an API key at verity.backworkai.com/keys.

Configuration

Claude Desktop

Edit ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "verity": {
      "command": "node",
      "args": ["/absolute/path/to/verity_mcp/build/index.js"],
      "env": {
        "VERITY_API_KEY": "vrt_live_YOUR_API_KEY"
      }
    }
  }
}

Claude Code

claude mcp add verity node /absolute/path/to/verity_mcp/build/index.js
export VERITY_API_KEY=vrt_live_YOUR_API_KEY

GitHub Copilot (VS Code)

Create or edit .vscode/mcp.json in your workspace (or open the user-level config via the command palette → MCP: Open User Configuration):

{
  "servers": {
    "verity": {
      "command": "node",
      "args": ["/absolute/path/to/verity_mcp/build/index.js"],
      "env": {
        "VERITY_API_KEY": "vrt_live_YOUR_API_KEY"
      }
    }
  }
}

OpenAI Codex

Via CLI (recommended):

codex mcp add verity --env VERITY_API_KEY=vrt_live_YOUR_API_KEY -- node /absolute/path/to/verity_mcp/build/index.js

Or edit ~/.codex/config.toml directly:

[mcp_servers.verity]
command = "node"
args = ["/absolute/path/to/verity_mcp/build/index.js"]
env = { VERITY_API_KEY = "vrt_live_YOUR_API_KEY" }

Available tools

Tool	What it does
`lookup_code`	CPT, HCPCS, ICD-10, or NDC lookup with RVU and coverage data
`search_policies`	Full-text search across LCDs, NCDs, and Articles
`get_policy`	Fetch full policy content by ID
`compare_policies`	Side-by-side coverage comparison across MAC jurisdictions
`get_policy_changes`	Policy updates since a given date or for a specific policy
`search_criteria`	Search extracted coverage criteria text by section
`list_jurisdictions`	MAC jurisdiction details and covered states
`check_prior_auth`	Prior auth requirements for procedure + diagnosis + state

Example prompts

Once connected, ask the assistant directly:

Is ultrasound guidance (76942) covered in Texas, and does it need prior auth?

What changed in Medicare CGM coverage policies in the last 30 days?

Compare coverage for J0585 (Botox) across JM and JH jurisdictions.

Show me the indications criteria for CPAP coverage.

Environment variables

Variable	Required	Description
`VERITY_API_KEY`	Yes	Your Verity API key (`vrt_live_…` or `vrt_test_…`)
`VERITY_API_BASE`	No	Override the API base URL (default: `https://verity.backworkai.com/api/v1`)

When to use MCP vs Agent Tools API

Use the MCP server when:

A developer or analyst is chatting with an AI assistant and needs ad-hoc coverage lookups
You want zero-code Verity access inside any AI client (Claude, Codex, Cursor, etc.)
The use case is exploratory — research, policy review, or answering one-off questions

Use the Agent Tools API when:

You're building a production application with an embedded agent
You need audit logging of every tool invocation
PHI must be explicitly rejected at the API layer
Your workflow requires human review checkpoints and approval flows
You need per-tool scope controls beyond a single API key

Last updated on

MCP Server

On this page