# MedPrice AI

> Hosted MCP server exposing US hospital procedure cost data to AI assistants

- name: io.github.medprice-ai/mcp-medprice-ai
- version: 0.0.1
- connection class: R0 (autonomous)
- trust: 87/100
- quality: 84/100 (good)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/io.github.medprice-ai%2Fmcp-medprice-ai
- tools: https://api.protogrid.dev/v1/servers/io.github.medprice-ai%2Fmcp-medprice-ai/tools

**An agent can connect right now, no human step.** Remote endpoint, no authentication, reachable on the last probe.

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "mcp-medprice-ai": {
      "type": "http",
      "url": "https://mcp.medprice.ai/mcp"
    }
  }
}
```


## Trust

- hygiene: 100
- liveness: 99
- freshness: 65
- provenance: 80
- drivers: +repository +namespace-matches-repo +reachable +uptime-97% ~updated-123d-ago
- Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed.

## Quality 84/100 (good)

### protocol: 88 (weight 25)

- pass `protocol.modern`: supports 2026-07-28 (server/discover)
- pass `protocol.stateless`: answers without a session
- pass `protocol.transport`: streamable HTTP
- warn `protocol.list_ttl`: tools/list declares ttlMs 0, so clients re-fetch it on every use

### authorization: n/a (weight 15)

- n/a `auth.prm`, `auth.as_metadata`, `auth.cimd`: no remote uses OAuth
- n/a `auth.secret_in_url`: no templated URL

### tool hygiene: 86 (weight 30)

- pass `tools.descriptions`: every tool has a description
- pass `tools.description_length`: descriptions are concise
- pass `tools.schemas`: every tool has an object input schema
- pass `tools.annotations`: 7 of 7 tools declare readOnlyHint or destructiveHint
- fail `tools.directory_hints`: no tool declares all four of readOnlyHint, destructiveHint, idempotentHint and openWorldHint; missing: batch_get_hospital_chargemaster_cost (idempotentHint); list_hospitals (idempotentHint); list_hospital_code_costs (idempotentHint); …
- pass `tools.token_cost`: about 3,029 tokens to load every tool
- pass `tools.api_dump`: tools are not a one-to-one API dump

### stability: 75 (weight 15)

- warn `stability.changes`: 12 tool changes in 30 days
- pass `stability.rug_pull`: no tool changed its meaning under the same name

### dependencies: n/a (weight 15)

- n/a `deps.known_vulns`, `deps.mcp_sdk_version`, `deps.resolvable`: no npm or PyPI package

- tools: 7, about 3,029 tokens to load them all
- tool set hash: a54d0ea0cc66e6d09f4a53cc943fc149
- tools last changed: 2026-10-20T19:51:17.822Z
- badge: [![protogrid quality](https://protogrid.dev/badge/io.github.medprice-ai/mcp-medprice-ai.svg)](https://protogrid.dev/servers/io.github.medprice-ai/mcp-medprice-ai)
- Checks run on what our credential-free, read-only probes observe; tools are never called and no code audit is performed.

### Tool changes

- 2026-10-20 `get_code_price_distribution` added
  - Returns the distribution of every hospital's price for a code_type/code, computed server-side: min/p10/q1/median/q3/p90/max, standard deviation, a histogram, and the real hospital nearest each percentile. Use this instead of paging through list_hospital_code_costs when you want to know how prices spread across hospitals (e.g. "is $12k high for MS-DRG 788?" - pass reference_price to get the number of hospitals priced below it). All figures are in terms of the chosen statistic (default: each hospital's average price).
- 2026-10-15 `list_codes` changed (inputSchema, outputSchema)
- 2026-10-09 `list_codes` changed (outputSchema)
- 2026-10-09 `list_hospital_code_costs` changed (outputSchema)
- 2026-10-09 `batch_get_hospital_chargemaster_cost` changed (outputSchema)
- 2026-10-09 `get_hospital_chargemaster_cost` changed (outputSchema)
- 2026-10-07 `list_codes` changed (description, inputSchema, outputSchema)
  - before: Returns every distinct code catalogued under a given code_type, paginated, with each code's raw chargemaster description and the number of hospitals reporting it. Use this to discover which codes exist under a code system (e.g. all CPT codes) before looking up prices with get_hospital_chargemaster_cost or list_hospital_code_costs.
  - after: Returns every distinct code catalogued under a given code_type, paginated, with each code's raw chargemaster description and the number of hospitals reporting it. Use this to discover which codes exist under a code system (e.g. all CPT codes) before looking up prices with get_hospital_chargemaster_cost or list_hospital_code_costs. Pass query to text-search codes/descriptions instead of listing a whole code_type. Pass hospital_id to restrict results to codes that one hospital actually reports.
- 2026-10-07 `list_hospital_code_costs` changed (inputSchema)
- 2026-10-07 `batch_get_hospital_chargemaster_cost` added
  - Looks up cost stats for several (code_type, code) pairs at one hospital (identified by hospital_id) in a single call - the batch counterpart to get_hospital_chargemaster_cost. Use this instead of calling get_hospital_chargemaster_cost once per code when pricing a fixed set of codes at one hospital (e.g. cross-linking a curated list of procedures on a hospital's page). Capped at 100 codes per call; a request over that is rejected with INVALID_ARGUMENT rather than silently truncated.
- 2026-09-28 `list_codes` changed (inputSchema, outputSchema)

Full history: https://api.protogrid.dev/v1/servers/io.github.medprice-ai%2Fmcp-medprice-ai/changes

## Remotes

- https://mcp.medprice.ai/mcp (streamable-http, auth none, reachable true, uptime30d 0.9745763)

## Packages

_none_

## Tools (7)

- `batch_get_hospital_chargemaster_cost`: Looks up cost stats for several (code_type, code) pairs at one hospital (identified by hospital_id) in a single call - the batch counterpart to get_hospital_chargemaster_cost. Use this instead of calling get_hospital_chargemaster_cost once per code when pricing a fixed set of codes at one hospital (e.g. cross-linking a curated list of procedures on a hospital's page). Capped at 100 codes per call; a request over that is rejected with INVALID_ARGUMENT rather than silently truncated.
- `get_code_price_distribution`: Returns the distribution of every hospital's price for a code_type/code, computed server-side: min/p10/q1/median/q3/p90/max, standard deviation, a histogram, and the real hospital nearest each percentile. Use this instead of paging through list_hospital_code_costs when you want to know how prices spread across hospitals (e.g. "is $12k high for MS-DRG 788?" - pass reference_price to get the number of hospitals priced below it). All figures are in terms of the chosen statistic (default: each hospital's average price).
- `get_hospital_chargemaster_cost`: Lookup hospital chargemaster cost
- `list_codes`: Returns every distinct code catalogued under a given code_type, paginated, with each code's raw chargemaster description and the number of hospitals reporting it. Use this to discover which codes exist under a code system (e.g. all CPT codes) before looking up prices with get_hospital_chargemaster_cost or list_hospital_code_costs. Pass query to text-search codes/descriptions instead of listing a whole code_type. Pass hospital_id to restrict results to codes that one hospital actually reports.
- `list_code_types`: Returns every distinct code_type (e.g. CPT, MS-DRG) with catalogued cost data, along with each type's distinct code count and total hospital reports. Unpaginated. Use this to discover which code systems have data before drilling into list_codes.
- `list_hospital_code_costs`: Returns cost stats for every hospital with a matching chargemaster entry for a code_type/code. Use this instead of calling get_hospital_chargemaster_cost once per hospital when comparing prices for the same procedure across hospitals (e.g. "which hospital has the cheapest MS-DRG 652?"). Defaults to returning every matching hospital (currently at most a few hundred) in one call. If the response's next_page_token is non-empty, the result set was truncated: call this tool again passing that exact value as page_token to get the next page, and keep doing so until next_page_token is empty - do not stop after one page and conclude a hospital has no data for this code. The response's total_count field (total across all pages) can be compared against how many results you've accumulated so far as a completeness check.
- `list_hospitals`: Returns the hospitals supported by the medprice.ai API, with their hospital_id (opaque handle), EIN, name, structured_locations (addresses with geocoded coordinates where available), last_updated_on, and revision history (with per-revision has_payer_data, revision_id, and added_at - when medprice itself ingested that revision, as opposed to revision_date's source-file-reported update date). Defaults to returning the entire registry (currently a few hundred hospitals) in one call. If the response's next_page_token is non-empty, the result set was truncated: call this tool again passing that exact value as page_token to get the next page, and keep doing so until next_page_token is empty - do not stop after one page and conclude the list is complete. The response's total_count field (total across all pages) can be compared against how many hospitals you've accumulated so far as a completeness check, e.g. before answering questions like 'does medprice.ai cover any hospitals in Iowa'.