Skills Artificial Intelligence Deploying OpenMed MCP Server Tools

Deploying OpenMed MCP Server Tools

v20260803
deploying-openmed-mcp
This skill deploys the OpenMed Model Context Protocol (MCP) server, exposing critical clinical NLP tools such as NER, PII extraction, and de-identification. It enables AI coding agents and chat clients to interact with OpenMed's capabilities via simple tool calls, running locally on-device to ensure user data privacy and security. Supports both stdio and Streamable HTTP deployment modes.
Get Skill
235 downloads
Overview

Deploying the OpenMed MCP server

openmed.mcp.server exposes OpenMed's clinical NLP as Model Context Protocol
tools, so coding agents (Claude Code, Codex) and chat clients can de-identify and
analyze clinical text by calling tools instead of writing glue code. It runs
on-device — models are local, no telemetry — and the server instructs
clients to send real PHI only to instances the user operates.

When to use this skill

When an agent or LLM client should be able to invoke OpenMed: add it to a
coding agent's MCP config, give a chat client de-id/NER tools, or run a shared
MCP endpoint for a team. For programmatic HTTP from your own services, prefer
serving-openmed-rest-api; for corpora, batch-processing-clinical-text.

Quick start

pip install "openmed[mcp]"                 # FastMCP / MCP SDK

# stdio transport (what coding agents spawn): default
python -m openmed.mcp.server

# Streamable HTTP transport (network-reachable):
python -m openmed.mcp.server --transport streamable-http --host 127.0.0.1 --port 8081
# Or embed it:
from openmed.mcp.server import create_mcp_server
server = create_mcp_server()              # FastMCP("OpenMed", ...) with tools+resources+prompts
server.run(transport="stdio")             # or "streamable-http"

CLI flags (build_arg_parser): --transport {stdio,streamable-http,http},
--host, --port, --streamable-http-path (default /mcp), --version.
Env equivalents: OPENMED_MCP_TRANSPORT, OPENMED_MCP_HOST,
OPENMED_MCP_PORT (8081), OPENMED_MCP_PATH.

The 7 tools (confirmed in openmed/mcp/server.py)

ToolWhat it doesKey args
openmed_analyze_textclinical NERtext, model_name (disease_detection_superclinical), confidence_threshold, group_entities, aggregation_strategy, sentence_*, keep_alive
openmed_extract_piidetect PII/PHI spanstext, model_name (default PII model), confidence_threshold (0.5), use_smart_merging, lang, normalize_accents
openmed_deidentifymask/remove/replace/hash/shift datestext, method (mask), confidence_threshold (0.7), keep_year, shift_dates, date_shift_days, keep_mapping, lang
openmed_list_modelslist registry modelscategory, pii_language, limit
openmed_list_pii_languagessupported PII languages + default models—
openmed_loaded_modelsresident-model status of the MCP runtime—
openmed_unload_modelfree one model or all inactive modelsmodel_name, all_models

It also registers resources — openmed://models, openmed://pii-languages,
openmed://examples (synthetic) — and prompts openmed-clinical-ner and
openmed-pii-deidentify that nudge the agent toward safe, correct calls.

Adding it to a coding agent

// Claude Code: .mcp.json (or ~/.claude.json) — stdio transport
{
  "mcpServers": {
    "openmed": {
      "command": "python",
      "args": ["-m", "openmed.mcp.server"],
      "env": { "OPENMED_PROFILE": "prod" }
    }
  }
}

For a shared HTTP deployment, run --transport streamable-http and point the
client at http://<host>:8081/mcp. The agent then sees the 7 tools and can call
e.g. openmed_deidentify on a snippet before sending it elsewhere.

Runtime config

The MCP server shares OpenMed's ServiceRuntime (ServiceRuntime.from_env()),
so the same env vars as the REST service apply: OPENMED_PROFILE,
OPENMED_SERVICE_PRELOAD_MODELS, OPENMED_SERVICE_KEEP_ALIVE,
OPENMED_SERVICE_MAX_RESIDENT_MODELS. Preload to avoid first-call latency;
openmed_unload_model/openmed_loaded_models let an agent manage memory.

Running in Docker

FROM python:3.11-slim
RUN pip install --no-cache-dir "openmed[mcp]"
ENV OPENMED_MCP_TRANSPORT=streamable-http \
    OPENMED_MCP_HOST=0.0.0.0 OPENMED_MCP_PORT=8081 \
    OPENMED_SERVICE_PRELOAD_MODELS="OpenMed/OpenMed-PII-SuperClinical-Small-44M-v1"
EXPOSE 8081
CMD ["python", "-m", "openmed.mcp.server"]

stdio servers are spawned by the client and don't need a port; use HTTP only for
shared/remote access, behind your own auth proxy. Mount the model cache so the
container starts offline.

Workflow

  1. Install + launch. pip install "openmed[mcp]", then

python -m openmed.mcp.server (stdio) or --transport streamable-http
for a shared endpoint.

  1. Configure the runtime via the ServiceRuntime env vars (profile,

preload, keep-alive, max resident) so first calls aren't cold.

  1. Register with the client. Add the mcpServers entry (stdio command, or

HTTP URL) to the agent's config; the 7 tools, resources, and prompts appear.

  1. Front HTTP with auth/TLS if remote — the server has none built in; keep

stdio/local for untrusted-network scenarios.

  1. Let the agent call tools (openmed_deidentify before sharing a snippet,

openmed_analyze_text for NER), and discover models via
openmed_list_models rather than hardcoding.

  1. Manage memory with openmed_loaded_models / openmed_unload_model.

Hand-off to / from OpenMed

  • Same engine: each tool calls openmed.analyze_text / extract_pii /

deidentify through the shared runtime — identical results to the library and
the REST service.

  • REST sibling: serving-openmed-rest-api exposes the same operations as

HTTP routes for non-agent callers.

  • Discovery: openmed_list_models / openmed_list_pii_languages mirror the

library's list_* functions — agents should query, not hardcode.

Edge cases & gotchas

  • stdio vs HTTP. Coding agents spawn the server over stdio (default) and

manage its lifecycle; use streamable-http only for a shared endpoint, and
put auth/TLS in front of it (the server has none built in).

  • PHI trust boundary. The server's instructions tell clients to send real

PHI only to instances the user controls. Keep it local/self-hosted; don't
point agents at an OpenMed MCP you don't operate.

  • keep_mapping=True returns a re-identification map in the

openmed_deidentify response — only enable for trusted agents, treat the
mapping as PHI, never log it.

  • No raw PHI in logs. Don't add transcript/body logging around the server.
  • Use synthetic examples in docs/tests/prompts — the bundled

openmed://examples resource is synthetic on purpose.

  • --transport http is accepted as an alias for streamable-http.

Standards & references

  • Model Context Protocol specification: https://modelcontextprotocol.io/
  • MCP transports (stdio, Streamable HTTP):

https://modelcontextprotocol.io/docs/concepts/transports

  • Claude Code MCP configuration:

https://docs.anthropic.com/en/docs/claude-code/mcp

  • OpenMed source: openmed/mcp/server.py (create_mcp_server, the 7 tools,

resources, prompts, main/build_arg_parser).

Info
Name deploying-openmed-mcp
Version v20260803
Size 7.62KB
Updated At 2026-08-04
Language