Back to MCP Servers

Ketryx MCP Server

Connects AI coding assistants to Ketryx compliance and quality management data for regulated product development in medical devices, digital health, and robotics.

Healthcare & Life Sciences by Ketryx OAuth2 active
Overview

Ketryx MCP is a provider-hosted Model Context Protocol server that brings live compliance intelligence from the Ketryx platform into AI workflows. It enables AI assistants like Claude, ChatGPT, Cursor, and GitHub Copilot to securely access an organization's quality management system (QMS), traceability matrix, design controls, and release data so AI-generated specs, test cases, and code stay aligned with regulatory requirements such as FDA QMSR, IEC 62304, ISO 13485, and ISO 14971.

The server unifies fragmented product lifecycle management (PLM), application lifecycle management (ALM), and source-control data into a single knowledge graph. Teams can query compliance status, identify traceability gaps, verify release readiness, and assess change impact using natural language, all without leaving their IDE, terminal, or CI/CD pipeline. Ketryx grounds responses in each organization's validated procedures using a neuro-symbolic approach that pairs LLM reasoning with a deterministic rules layer, producing auditable answers tied to the team's actual quality system.

Announced in beta on March 31, 2026, the server is available through a private access program. Access, server URL, and detailed tool schemas are provisioned to approved beta participants. The Ketryx organization maintains other public repos (such as ketryx-github-action), but the MCP server itself is hosted by Ketryx rather than distributed as an open-source package.

Setup Guide

Ketryx MCP is in private beta. There is no public npm/PyPI package or GitHub repo for the server itself; access, the server URL, and credentials are issued by Ketryx after approval.

Steps:

  • Request beta access at https://www.ketryx.com/mcp-access-request. Beta is open to organizations developing regulated products in medical technology, digital health, and robotics.
  • Once approved, Ketryx provides a server URL and authentication credentials tied to your Ketryx workspace.
  • Configure your MCP client (Claude Desktop, Cursor, Copilot, ChatGPT) to point at the Ketryx-provided endpoint.

Example Claude Desktop config (placeholder, use the values Ketryx gives you):

{
  "mcpServers": {
    "ketryx": {
      "url": "https://<your-ketryx-mcp-endpoint>",
      "headers": {
        "Authorization": "Bearer <KETRYX_MCP_TOKEN>"
      }
    }
  }
}

Prerequisites:

  • An active Ketryx workspace with your QMS, traceability, and project data configured.
  • Beta program approval from Ketryx.
  • An MCP-capable AI client (Claude, ChatGPT, Cursor, or Copilot).
Use Cases
  • Ask an AI assistant whether a release candidate has full traceability from requirements through design, code, and tests before approving deployment.
  • Generate IEC 62304 or FDA pre-submission documentation drafts grounded in the organization's validated QMS procedures and Ketryx item data.
  • Identify traceability gaps (orphan requirements, untested risks, missing verification evidence) directly from the IDE.
  • Run automated change impact assessments when modifying a software component in a regulated medical device codebase.
  • Verify that AI-generated code, specs, or test cases comply with the team's documented quality system before merging.
Example Prompts
  • "Is release 2.4 of our infusion pump firmware ready to ship? List any traceability gaps or open risks."
  • "Show me all requirements that don't have linked verification tests in the current Ketryx project."
  • "Draft an IEC 62304 software development plan section for the alarm subsystem using our QMS templates."
  • "What's the change impact if I modify the dose calculation module? Which risks and tests are affected?"
  • "Summarize the current compliance status of our Class II device project, including any unresolved findings."
Pros
  • Official, vendor-hosted MCP server maintained by Ketryx, so compliance data stays in sync with the source system.
  • Grounds LLM responses in validated QMS procedures via a deterministic rules layer, producing auditable answers rather than free-form guesses.
  • Unifies PLM, ALM, and source-control data into a single knowledge graph queryable from any MCP-capable client.
  • Fits regulated development workflows (FDA QMSR, IEC 62304, ISO 13485, ISO 14971) without forcing teams off their existing tools.
Limitations
  • Private beta only as of May 2026; access requires Ketryx approval and a Ketryx workspace.
  • No public documentation of the exact tool list, server URL, or schema, so capabilities must be confirmed with Ketryx after onboarding.
  • Useful only to teams already using Ketryx as their compliance platform; there is no value as a standalone connector.
Alternatives
  • Greenlight Guru (medical device QMS) does not currently offer an MCP server but exposes a REST API that can be wrapped.
  • Jama Connect for ALM and traceability; no official MCP server, community connectors exist for its REST API.
  • Polarion ALM via custom MCP wrappers around its OpenAPI for teams that want similar traceability queries.