Medplum

# Fix list: Medplum

From Anchor Terminal's listing at https://www.anchorterminal.com/tools/medplum, the October 2026 research run, assessed 14 October 2026. Grade B, 68.7 out of 100.

This is everything the published grade says the listing lacks, the biggest possible gain to the total first. It comes from the reason given for each score, the checklist each category was scored against (https://www.anchorterminal.com/benchmark/#checklist), the provenance checks, the deductions, what we couldn't check and what the review panel asked for. A fix counts at the next check, once it's public.

For a coding agent working on Medplum: work through the items below in the product, its docs and its public pages. Each category gives the reason for its score, with the points each checklist item earned, and the checklist itself, so the gap is the items that earned less than their points. Change the product, not the wording, and keep a note of what you changed and where it's published.

## 1. Payments & pricing, 25 out of 100, up to 9.4 more on the total

Why it scored 25: Read with the hosted rubric, since the hosted service is what an agent would call. No x402, MPP or L402 (0). Plan prices are public, Production $2,000 and Premium $6,000 a month, with no per-call price (10). A free hosted plan exists for prototyping, and whether sign-up asks for a card was not established (15 of 20). A person registers at app.medplum.com and creates a ClientApplication in a browser (0).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-payments):

The published rubric, also on the [x402 page](https://www.anchorterminal.com/x402/).

- 40, a machine payment protocol (x402, MPP or L402) on the tool's own endpoints. 10 to 30 when it covers only some endpoints or only goes through a third party, and the note says which.
- 20, per-call or per-unit pricing published without a login. 10 for public plan-only pricing, 0 for "contact sales" or prices behind a login.
- 20, a free tier or trial that doesn't need a card.
- 20, autonomous onboarding, meaning an agent can get access without a person signing up in a browser (keyless use, x402, a programmatic key API).

Payment platforms and agent wallets rarely charge for their own API over a machine protocol, so the first line has steps for them, and the highest one that applies counts. 40 when x402, MPP or L402 runs on all their own endpoints, 30 when it runs on part of their own API, 25 when their merchants can accept one, 20 for running a facilitator, 15 for paying as a buyer, and 0 when the only protocol is their own. Merchant acceptance sits above a facilitator because the platform's own customers can charge agents through it, while a facilitator settles for sellers who wire up the protocol themselves. The counter-argument (a facilitator does more for the protocol as a whole) has a point. Each note says which step applied.

Open-source software you run yourself is scored on its hosted or paid option if it has one. A free, self-hosted package with nothing to buy gets 20, 20 and 20 for the last three lines, and 0 to 40 for the first only if it ships a payment protocol.

## 2. Reliability, 75 out of 100, up to 5 more on the total

Why it scored 75: Read with the hosted lines and scored on the hosted FHIR API at api.medplum.com, which is what an agent would call. The docs name a status page at status.medplum.com with incident history and root cause analyses (20). Its robots.txt disallows every path, so the 90-day record was not read (5). Rate limits are published per IP and tier, with a weighted FHIR interaction quota (15). Every limited response carries a `RateLimit` header with remaining units and reset time, the docs say not to retry before the reset, and conditional create and upsert make writes safe to repeat (15). The docs state a 99.99 per cent monthly SLA with service credits for the production hosted platform (10). The FHIR API is generally available (10).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-reliability):

Hosted APIs, MCP servers, models and platforms.

- 20, a public status page with component history (Statuspage, Instatus, BetterStack or the vendor's own).
- 0 to 30, the incident record for the last 90 days on that page. 30 for a clean record or trivial incidents only, 20 for minor incidents only, 10 for one major outage (an hour or more of a core API down, or errors across the board), 0 for several. 5 when there's no history we could read, and the note says so.
- 15, rate limits documented with numbers.
- 15, documented 429 or overload handling (Retry-After, backoff guidance), and idempotency keys or safe-retry guidance where writes are involved.
- 10, an SLA published for any paid tier.
- 10, the surface agents use is generally available, not beta or preview.

Local packages, SDKs, frameworks and stdio MCP servers.

- 20, installs from an official package with supported runtimes stated.
- 25, a public CI and test suite, passing on the default branch.
- 0 to 25, open crash or regression issues relative to activity (25 for few and handled, 0 for many, old and unanswered).
- 15, semver discipline and breaking changes called out in a changelog.
- 15, version 1.0 or later, or declared stable.

Protocols are read from their reference implementations, the public facilitators or servers, spec stability and test vectors.

## 3. Security & auth, 74 out of 100, up to 4.6 more on the total

Why it scored 74: OAuth 2.0 with client credentials, SMART scopes, client assertion, token exchange and mTLS. Client secrets belong to a ClientApplication that an admin can delete (28 of 30). AccessPolicy resources give read-only access by resource or element, IP access rules and per-member policies. No confirmation step for destructive writes, including through the MCP tool (15 of 20). Records are written by other users and returned to the model, and no prompt-injection guidance was found beyond a caution in the AI coding page (3 of 15). FHIR AuditEvent resources and log streaming on paid plans (12 of 15). security.txt valid with PGP, SOC 2 Type II, advisories published on GitHub, no bug bounty found (16 of 20).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-security):

- 0 to 30, the credential model. 30 for OAuth 2.1 with scopes, or scoped and revocable keys with rotation. 20 for plain revocable API keys. 10 for one all-powerful key. 10 off when a secret can travel in a URL query string as a documented option.
- 0 to 20, read-only or least-privilege modes, and confirmation or approval for destructive actions.
- 0 to 15, prompt-injection posture where the tool returns untrusted content (documented mitigations or guidance). A tool that returns no untrusted content gets 10.
- 0 to 15, audit logs or per-call visibility for the operator.
- 0 to 20, a security programme. security.txt or a disclosure policy, a bug bounty, SOC 2 or ISO 27001, advisories handled in public.

Models are read for retention, whether API data trains models (and whether that's off by default), zero-retention options and certifications. Frameworks for telemetry defaults, approval hooks, guardrails and sandboxing.

## 4. Agent ergonomics, 73 out of 100, up to 4.4 more on the total

Why it scored 73: FHIR `_count`, `_elements`, `_summary` and `_include` size responses (22 of 25). Cursor and offset paging, chained search and `_filter` (20). Errors are OperationOutcome with HTTP status, and 429s carry the `RateLimit` header (15 of 20). Conditional create and upsert give safe retries, but no idempotency key header and no MCP readOnlyHint or destructiveHint on `fhir-request` (8 of 20). Few required parameters, with the official SDK in TypeScript only and Python shown as raw requests (8 of 15).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-ergonomics):

- 0 to 25, context cost. For MCP, the number and size of the tool definitions (25 for ten or fewer compact tools, 15 for 11 to 30, 5 for more than 30, plus up to 10 back for toolsets, dynamic loading or read-only subsets). For APIs, whether responses can be sized (field selection, limits, summaries).
- 20, pagination, filtering and output-size controls.
- 20, actionable, documented error responses, codes and messages an agent can recover from.
- 20, idempotency or safe retries, and for MCP the `readOnlyHint` and `destructiveHint` annotations.
- 15, sensible defaults, few required parameters, and official SDKs in at least two languages.

Models are read for tool use, structured output, prompt caching, context length, batch and SDKs. Frameworks for how much code and how many defaults a tool-calling agent with MCP needs.

## 5. Schema & documentation, 82 out of 100, up to 2.9 more on the total

Why it scored 82: The server generates an OpenAPI 3.1 document (packages/server/src/openapi.ts) and publishes a FHIR CapabilityStatement. The live copies sit on api.medplum.com, which robots.txt disallows, so they were not read (20 of 25). llms.txt indexes 2,600 lines of docs (10). Each FHIR resource and operation has a reference page, and the MCP doc explains when to use each tool (15 of 20). FHIR resources are typed with required fields and value sets, though the MCP tool takes `method` and `path` as free strings and `body` as any JSON (12 of 15). Docs carry curl, TypeScript and Python examples, with errors as FHIR OperationOutcome (12 of 15). Semver and GitHub releases with notes (15). Total 84, less 2 for the loosely typed MCP schema in source, 82.

The checklist (https://www.anchorterminal.com/benchmark/#checklist-schema):

APIs and MCP servers.

- 25, a machine-readable contract (a public OpenAPI file or similar; for MCP, typed JSON Schema inputs on every tool).
- 10, llms.txt or Markdown docs served for agents.
- 0 to 20, descriptions that say what a tool is for, when to use it and when not to, read from the tool definitions in the source or the API reference.
- 0 to 15, typed inputs with enums, constraints and required fields, and no free-form JSON blobs.
- 0 to 15, examples and documented error responses.
- 15, versioning and a public changelog.

Models are read from the API reference, the OpenAPI file, llms.txt, the structured-output and tool-use docs and the model cards. Frameworks from docs a model can follow, typed interfaces, examples and the API reference.

## 6. Transparency & trust, 72 out of 100, up to 2.5 more on the total

Made of editorial 65, provenance 78.

Why it scored 72: Apache 2.0 source for the server (30). The privacy policy was last updated on 5 April 2022 and is written for licensed clinicians, the terms on 11 April 2023 grant Medplum a broad licence over user content, and PHI is handled under a BAA. No retention periods were found (12 of 30). A written version policy sets one year active and one year maintenance per major, with security patch deadlines (18 of 20). The docs say hosting runs on AWS across several Availability Zones. No sub-processor list was found (5 of 20).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-transparency):

- 0 to 30, source availability and licence clarity. 30 for open source under an OSI licence, 15 for closed with clear terms, 0 for unclear terms.
- 0 to 30, data handling and retention statements that agree with each other (privacy policy, DPA, retention periods, subprocessors).
- 0 to 20, a deprecation policy or notices with dates.
- 0 to 20, telemetry disclosed with an opt-out (local software), or subprocessors and data locations disclosed (hosted).

The other half of Transparency and trust is the provenance score, computed from checked facts (below). The category score is the mean of the two.

Provenance checks not met in full (half of this category, computed from checked facts):

- Domain age: medplum.com, no registry record we could read (0 of 15)
- Terms of service: read, states 6 of the 7 things a reader expects, and has 1 clause that costs points (7.1 of 10)
- Privacy policy: read, states 3 of the 8 things a reader expects (6.3 of 10)

## 7. Maintenance & community, 93 out of 100, up to 0.6 more on the total

Why it scored 93: v5.2.3 tagged on 8 October 2026 (30). Nine tags between 24 September and 8 October 2026 (20). 571 open issues, with commits landing daily and pull requests reviewed, and Discord plus email support. Reply times were not measured (18 of 25). Listed in the official MCP registry as com.medplum/mcp, and @medplum/core 5.2.3 is current with 269,965 weekly downloads (15). The Build workflow passed on main on 14 October 2026, with CodeQL and OpenSSF Scorecard runs (10).

The checklist (https://www.anchorterminal.com/benchmark/#checklist-maintenance):

- 0 to 30, time since the last release, or the last published model or API change for a closed service. 30 within 30 days, 20 within 90, 10 within 180, 0 older.
- 20, at least three releases or dated changelog entries in the last 90 days.
- 0 to 25, responsiveness. Issues and pull requests answered on GitHub (the open issues and how recent the replies are). For closed services, a public changelog and a support or community channel that answers, 0 to 15.
- 15, presence in the official MCP registry under a verified namespace (MCP servers), or current official SDKs (APIs and models).
- 10, package health, current dependencies and CI.

Models are read for deprecation notice periods and model churn rather than release counts.

## Deductions

Each comes off the total. A fixed and documented problem counts for less at the next check.

- 2026-07-24: GHSA-m44r-7c5h-m6mj, high severity. The external auth callback accepted redirect URIs that only had to start with a registered one, which could leak authorisation codes cross-origin. Patched in @medplum/core 5.1.6 and disclosed on GitHub (https://github.com/medplum/medplum/security/advisories/GHSA-m44r-7c5h-m6mj). With GHSA-ch8p-j6cm-r7w5 (2 June 2026, client secret exposure in self-hosted dynamic registration), both fixed and disclosed in public, so the deduction is small (-2).

## What we couldn't check

What we couldn't read counted as absent. Publishing it on a page a plain HTTP fetch can read (not only in a browser) lets the next check count it.

- unchecked: the incident history at status.medplum.com. Its robots.txt disallows all paths, so it was not fetched
- unchecked: the live OpenAPI document and CapabilityStatement on api.medplum.com, whose robots.txt disallows all paths
- unchecked: whether the free hosted plan asks for a card at sign-up
- unchecked: the domain registration date for medplum.com
- GHSA-fjgc-c3pj-xx2c is cited in the MCP source as the reason for an SSRF guard but was not found among the repository's published advisories

## Weaknesses

- The MCP `fhir-request` tool can delete or overwrite any resource the token allows, with no readOnlyHint or destructiveHint
- The privacy policy was last updated on 5 April 2022 and the terms on 11 April 2023
- No sub-processor list or retention periods were found
- Two high-severity auth advisories were published and patched in 2026
- Official SDK in TypeScript only

## What costs an agent a turn today

The notes we give agents before they call it. Each one is a workaround an agent shouldn't need.

- Use a ClientApplication with an AccessPolicy limited to the resources the task needs, and set `readonly` where no writes are needed
- Read the `RateLimit` header and wait for the reset time after a 429. Writes cost 100 quota points, searches 20 and reads 1
- Create patients with conditional create (`If-None-Exist`) or upsert on an identifier so a retry doesn't make a duplicate
- Pass paths to `fhir-request` relative to the FHIR R4 base, with no leading slash or `/fhir/R4` prefix
- Use `Prefer: respond-async` for large bundles. Operations inside the background job don't count against the quota

## When it's done

Send what changed and where it's published as a dispute (https://www.anchorterminal.com/builders/#disputes, or `POST https://www.anchorterminal.com/api/v1/contact` with `"kind": "dispute"`). Disputes are answered in public, and the listing is checked again by the same checklist. Paying for an audit or a listing claim changes nothing here.