Skip to main content
Glama
com.blackswancausallabs/openfda-mcp

com.blackswancausallabs/openfda-mcp

Official

openFDA MCP

An MCP server exposing FDA regulatory metadata (drugs, biologics, and medical devices) through the openFDA API.

Built by Black Swan Causal Labs as the identifier-resolution layer for a real-world-evidence (RWE) case roster: given an FDA application number or a product name, resolve it to authoritative regulatory metadata.

Why this exists

There are other openFDA MCP servers, and several are broader. This one is narrow on purpose: it is the instrument that resolved the application numbers in a specific published RWE dataset, and it exists so that dataset can name the tool that produced it.

That matters more than it might sound. Whether BLA 125123/2058 resolves to a particular product, or DEN160026 to a particular device class, is a decision made by a piece of software — and a different wrapper can yield a different roster. "We used openFDA" is not a sufficient methods statement; "openfda-mcp v0.1.0" is. If you use this in research, pin the version.

Its practical edge over a general openFDA client is the device half: resolving a CDRH submission number to a risk class takes a three-hop chain (number → product code → classification) with two non-obvious traps, both handled here.

Related MCP server: openfda-mcp-server

Tools

Drugs and biologics (/drug/* — CDER, CBER)

Tool

Purpose

search_drug_label

Search SPL label text, optionally scoped to a section

lookup_drugsfda_application

Drugs@FDA record for an NDA/BLA/ANDA number

resolve_drug_to_application

Brand or generic name → application number(s)

screen_for_rwe_signals

Experimental. Sweep labels for RWE signals

Devices (/device/* — CDRH)

Tool

Purpose

lookup_device_submission

K / DEN / P / H number → device record

classify_device_product_code

Product code → device class + medical specialty

validate_device_application

Full chain: number → class, specialty, category

Install

pip install openfda-mcp

Add to your MCP client config:

{
  "mcpServers": {
    "openfda": {
      "command": "openfda-mcp",
      "env": { "OPENFDA_API_KEY": "${OPENFDA_API_KEY:-}" }
    }
  }
}

The API key is optional. Without one, openFDA allows 40 requests/min and 1,000/day, which is enough for interactive use. A free key raises it to 240/min and 120,000/day — worth having for bulk sweeps.

Two findings worth knowing

Both were established empirically and are not obvious from FDA's docs.

De Novo grants live in the 510(k) endpoint. DEN###### numbers are stored in the k_number field of /device/510k. There is no De Novo endpoint, and looking for one leads to the wrong conclusion that De Novo numbers can't be resolved. They can.

HDE numbers are not in openFDA at all. Neither the 510(k) nor the PMA endpoint carries H######, so no product code — and therefore no classification — is retrievable. This server still reports device_class: "III" for them, by regulatory inference: HDE is by definition the pathway for devices that would otherwise require a PMA. medical_specialty stays null, because that one really is unavailable, and device_class_source says which is which.

Transient failures are never silent

A genuine absence and a failed request are different things, and this package keeps them different:

  • not found (HTTP 404, or 200 with no results) → returns None; safe to cache

  • transient failure (timeout, connection error, 429, 5xx) → retried with backoff, then raises OpenFDATransientError; never cache this

  • rejected request (other 4xx) → raises OpenFDARequestError

This is a direct response to a real defect: an earlier version swallowed every exception and returned None, so a single read timeout on one application number was cached as a real miss and silently blanked two fields on that record for weeks. Cached failures are indistinguishable from real absences, which makes them the worst kind of silent data loss.

On screen_for_rwe_signals

It is unvalidated. There is no ground-truth oracle for a discovery sweep, and below the strongest hits the results are dominated by applications whose labels use "registry" in an unrelated sense. Treat its output as candidates for human review — not as a finding, and not as a count to report. Establishing recall against a held-out set of known cases is open work.

Development

pip install -e ".[dev]"
pytest              # unit tests, offline
pytest -m live      # live checks against api.fda.gov

Live tests assert against known-good fixtures (K203571 → class II Ophthalmic, DEN160026 → class II Immunology, BLA761180 → LEO Pharma) so a change on FDA's side surfaces as a test failure rather than as quietly wrong data.

License

MIT

Available Tools

7 tools
classify_device_product_codeB

Resolve a CDRH product code to its device class and medical specialty.

Inputs: product_code (required). Returns: object

lookup_device_submissionA

Look up a CDRH device submission by its number.

Handles all four device pathways: 510(k) K-numbers ("K203571"), De Novo grants ("DEN160026" — stored in the 510(k) endpoint, not a separate one), PMA P-numbers ("P230044"), and HDE H-numbers. Supplement suffixes ("P230044/S001") are stripped before lookup.

Inputs: number (required). Returns: object

lookup_drugsfda_applicationB

Fetch Drugs@FDA application metadata for an NDA/BLA/ANDA number.

Inputs: application_number (required). Returns: object

resolve_drug_to_applicationA

Resolve a brand or generic drug name to its NDA/BLA application number(s).

Bridges drug-name-only sources (registry inventories, product lists) to FDA regulatory metadata. Searches the brand and generic name fields directly rather than label prose, so another product merely mentioning this drug does not produce a false match.

Inputs: drug_name (required), limit. Returns: object

screen_for_rwe_signalsA

EXPERIMENTAL. Sweep drug labels for registry / real-world-evidence signals.

Searches the label corpus for terms suggesting registry, natural-history or other real-world evidence, then aggregates hits by application number with snippets showing where each term matched.

This tool is unvalidated and has a known false-positive problem. It has no ground-truth oracle, and below the strongest hits the results are dominated by applications whose labels use "registry" in an unrelated sense. Its output is a candidate list for human review — not a finding, and not a count you should report.

Inputs: extra_terms, limit_per_term. Returns: result

search_drug_labelB

Search openFDA drug labels (SPL) for a term.

Inputs: term (required), skip, limit, section. Returns: object

validate_device_applicationA

Validate a device number and derive class, specialty and category in one call.

Chains lookup_device_submission -> classify_device_product_code, which is the full three-hop resolution: number -> product_code -> classification.

Inputs: number (required). Returns: object

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.9/5.0

    Scored across 7 tools

    Disambiguation5/5

    Each tool targets a distinct resource/action: drug label search, application metadata lookup, name-to-application resolution, RWE signal screening, device submission lookup, product code classification, and full validation. The drug and device groups are clearly separated, and descriptions clarify any potential overlap (e.g., validate_device_application explicitly chains the other two device tools).

    Naming Consistency4/5

    All tool names follow a verb-first, snake_case convention (search, lookup, resolve, screen, classify, validate), but the noun phrases vary in structure: some are direct objects (drug_label, device_submission), some use prepositions (drug_to_application, for_rwe_signals). This is mostly consistent and readable, with minor deviations.

    Tool Count5/5

    Seven tools is well-scoped for covering both drug-related (labels, applications, name resolution, RWE signals) and device-related (submissions, product codes, validation) workflows. Each tool serves a clear purpose without bloat.

    Completeness4/5

    The core workflows are covered: drug label search, drug name-to-application resolution, application metadata retrieval, RWE signal screening, device submission lookup, product code classification, and full validation. Minor gaps exist, such as no device search by name and no adverse event data, but the primary lifecycle is well supported.

    Maintenance

    ActivityInactive
    ResponsivenessNo recent community issues

    Related MCP Connectors

    Related MCP Servers

    • A
      license
      A
      quality
      D
      maintenance
      Enables LLMs to search FDA drug labels and adverse event data via the OpenFDA API, supporting natural language queries for drug safety information.
      2
      MIT
    • A
      license
      A
      quality
      D
      maintenance
      Provides LLM-optimized access to FDA data through the OpenFDA API and direct regulatory decision document retrieval, covering all 21 OpenFDA endpoints plus 510(k) summaries, De Novo decisions, and PMA approvals.
      4
      1
      MIT
    • A
      license
      Not graded
      quality
      B
      maintenance
      Enables querying FDA medical device regulatory databases including 510(k) clearances, PMA approvals, recalls, and adverse events through natural language.
      102 npm
      MIT