mcpField
P
Pa Medical Exclusions

Pa Medical Exclusions

Remote
pipeworx.ioLegal & Government0Updated 2w ago

About Pa Medical Exclusions

Pa Medical Exclusions is an MCP server for checking Pennsylvania’s Medicheck precluded-provider list. It supports searches by name and provider identifiers, while distinguishing current preclusion status from records that are unknown or no longer active.

Key features

  1. 01Screen by name, NPI, license, or FEIN
  2. 02Match identifiers with defined normalization
  3. 03Return DHS status and preclusion flag
  4. 04Summarize statuses and effective-date ranges
  5. 05Search the historical preclusion list

Use cases

  1. 01Check a provider before Pennsylvania Medicaid participation
  2. 02Verify a provider using NPI or license
  3. 03Review past preclusions by effective date
  4. 04Separate unknown statuses from recognized ones

MCP support

Exposes provider screening and list-coverage tools to MCP clients. It can run as a standalone stdio MCP server or be accessed through the Pipeworx gateway.

Configuration

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

Endpointhttps://gateway.pipeworx.io/pa-medical-exclusions/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

@pipeworx/pa-medical-exclusions

Pennsylvania Department of Human Services (DHS) Medicheck Precluded Providers List screening — check a provider by name, NPI, license number, or FEIN against the ~6,750-row list of providers and businesses ever precluded from Pennsylvania's Medical Assistance (Medicaid) program, the state-level counterpart to the federal HHS OIG LEIE (leie pack) and the sibling of ny-omig-exclusions (New York), ca-medical-exclusions (California), tx-medical-exclusions (Texas) and fl-medical-exclusions (Florida).

Part of Pipeworx — an MCP gateway connecting AI agents to 1831+ live data sources. This is an independent, unofficial integration — not affiliated with, endorsed by, or published by the upstream provider.

Tools

  • pa_medical_check_exclusion(name?, npi?, license?, fein?, limit?) — screens a provider against the DHS Medicheck list. An NPI, license, or FEIN match is an identification (all three are exact identifiers); a name match — even an exact one — is a candidate lead only, since the list carries no date of birth to disambiguate same-named providers. Every match carries DHS's own status field and a currently_precluded flag derived from it — null with status_unknown: true on the minority of rows where DHS's own status is blank, rather than a guess.
  • pa_medical_exclusion_coverage() — total rows ever recorded, how many are currently precluded vs. ended vs. status-unknown, how many carry an NPI, license number, or FEIN, the distinct status values DHS uses, and the oldest/newest preclusion effective date. Pass year or start_date/end_date to instead count preclusions whose effective date falls in that range, over the whole list.

Auth

Keyless.

Data sources

  • Pennsylvania Department of Human Services (DHS), Medicheck Precluded Providers List — the state's published roster of individuals, businesses and other entities precluded from participating in Pennsylvania's Medical Assistance program, searchable at https://www.humanservices.dhs.pa.gov/Medchk/ and downloadable as a full text file directly from that same application.
Why no open-data-portal mirror — checked, not assumed

A Socrata catalog search of data.pa.gov (api.us.socrata.com/api/catalog/v1?domains=data.pa.gov) for "medicheck", "precluded", "excluded provider" and "medicaid sanction" (2026-10-17) returned zero matching datasets. DHS's own Medicheck search application is the only source.

Why this is baked, not a live proxy

Per root CLAUDE.md's standing rule for a table this size, the full list (6,751 rows at capture time, ~2.7MB as generated TypeScript) is baked by scripts/bake-index.mjs into src/pa-medchk-index-data.ts, registered in workers/gateway/src/pack-baked-indexes.json, uploaded to KV at deploy, and injected into every call as args._bakedIndex — never a static module-scope import (fleet #2754: six packs doing that put ~48MB of retained heap into every gateway isolate and cost about one call in four as a Cloudflare 1102). A missing or malformed injection throws loudly rather than answering "not found" — see src/index.test.ts.

Unlike tx-medical-exclusions, no ASP.NET VIEWSTATE postback is needed: the Medicheck application's "Download Text File" button is a plain, unauthenticated GET, verified live 2026-10-17.

Re-run node mcps/pa-medical-exclusions/scripts/bake-index.mjs periodically and recommit; data_as_of on every response says how stale the baked copy is.

Status is messy — currently_precluded is derived and labelled

DHS's own Status field (Precluded, Terminated, Reinstated, Suspended, Withdrawal, Voluntary Withdrawal, Permanent Preclusion, Permanent Withdrawal, Terminated-No Cause) is the closest thing to a current-state flag, but 466 of 6,751 rows (2026-10-17 capture) carry a blank Status — checked directly, not assumed: EndDate presence does not cleanly correlate with Status either way (plenty of Precluded rows also carry an EndDate), so EndDate cannot be used to resolve the blanks. This pack derives currently_precluded from Status alone when Status is one of DHS's own recognized values, and reports currently_precluded: null with status_unknown: true on the blank rows — a guess here would be the silent wrong-answer shape root CLAUDE.md's loud-failure rule exists to prevent.

Ever-precluded vs. currently-precluded

Same shape as the four precedent state packs: this download is the ever precluded list, not a currently-precluded-only list — rows go back to 1957 and include providers long since reinstated, terminated, or withdrawn. ever_vs_current_caveat states this on every response.

Matching

An npi query matches only exact 10-digit numeric NPI values. A license query matches LicenseNumber, letters/digits only (case-insensitive, punctuation/whitespace stripped) — Pennsylvania license numbers are alphanumeric (e.g. AJ006399L) and sometimes carry an out-of-state annotation in parentheses (e.g. 172434 (NY)), so license: "172434" matches that row too. A fein query matches the federal EIN exactly, digits only. Name queries match against the provider's assembled individual name (first/middle/last), business name, or DHS's recorded alias, any word order, case/punctuation-insensitive.

Reachability

Verified live 2026-10-17 with a laptop curl against https://www.humanservices.dhs.pa.gov/Medchk/MedchkSearch/Medchk?target=_blank: 200, text/plain, quoted comma-delimited CRLF file, 6,751 data rows. No session state, cookies, or form replay required — simpler than the VIEWSTATE handshake tx-medical-exclusions needs.

Quick Start

Add to your MCP client (Claude Desktop, Cursor, Windsurf, etc.):

{
  "mcpServers": {
    "pa-medical-exclusions": {
      "url": "https://gateway.pipeworx.io/pa-medical-exclusions/mcp"
    }
  }
}
What this endpoint actually serves

tools/list at https://gateway.pipeworx.io/pa-medical-exclusions/mcp returns the tools in the table above plus the shared Pipeworx meta-tools — ask_pipeworx, discover_tools, search_within, remember/recall and the rest of the gateway-wide set. So the tool count you see is larger than this table: a single-pack endpoint currently lists roughly 30 shared tools alongside the pack's own. The connection's initialize response states its exact scope, and is the authoritative answer for a given day.

This is deliberate, not multiplexing by accident. The meta-tools are what let a scoped connection answer a question this pack does not cover — via ask_pipeworx, which routes across the whole catalog — without you adding a second MCP server. There is currently no way to mount a pack endpoint without them; if the extra schemas cost you more context than the routing is worth, connect to the full gateway once rather than to several pack endpoints.

Or connect to the full Pipeworx gateway to get every pack's tools listed directly, instead of just this one's:

{
  "mcpServers": {
    "pipeworx": {
      "url": "https://gateway.pipeworx.io/mcp"
    }
  }
}

Both URLs reach the same gateway and the same 1831+ data sources. The only difference is which pack's tools are listed directly; ask_pipeworx reaches all of them from either one.

No MCP client? Call it over HTTP

curl -X POST https://gateway.pipeworx.io/v1/tools/pa_medical_check_exclusion \
  -H 'Content-Type: application/json' \
  -d '{"name":"Zeidler"}'

No account needed for the first calls. Inspect any tool: GET https://gateway.pipeworx.io/v1/tools/pa_medical_check_exclusion. Find one: POST https://gateway.pipeworx.io/v1/tools/search_packs with {"query":"..."}.

Standalone (no gateway account)

This package also runs as a local stdio MCP server — no Pipeworx account, no gateway round-trip:

{
  "mcpServers": {
    "pa-medical-exclusions": {
      "command": "npx",
      "args": ["-y", "@pipeworx/mcp-pa-medical-exclusions"]
    }
  }
}

Or run it directly to confirm it starts:

npx -y @pipeworx/mcp-pa-medical-exclusions

It speaks MCP over stdin/stdout and answers initialize/tools/list/tools/call for only this pack's tools — none of the shared meta-tools the gateway connection above adds. Same source, same tools, no ask_pipeworx routing.

Using with ask_pipeworx

Instead of calling tools directly, you can ask questions in plain English — this works on the pack endpoint above as well as on the full gateway:

ask_pipeworx({ question: "your question about Pa Medical Exclusions data" })

The gateway picks the right tool and fills the arguments automatically.

More

License

MIT

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.