🏥 Healthcare & FDA Intelligence MCP

One MCP connection. The FDA, NIH, CMS, CDC, and ClinicalTrials.gov — all agent-ready.

Healthcare & FDA Intelligence MCP is a Model Context Protocol server that gives any AI agent — Claude, ChatGPT, Cursor, or a custom pipeline — direct, structured access to the core US healthcare and life-sciences data streams: drug labels and approvals, adverse events, recalls, warning letters, drug shortages, the Orange and Purple Books, 510(k) device clearances, pharma catalysts, clinical trials, NIH grants, CMS Open Payments, PubMed, and CDC surveillance.

Connect once and call fifteen clean tools. Each returns normalized JSON your agent can reason over immediately — no openFDA pagination, no ClinicalTrials.gov query syntax, no Sunshine Act CSV wrangling.

Why this exists

Healthcare data is the richest free public dataset in existence and one of the most fragmented. openFDA, ClinicalTrials.gov, NIH RePORTER, CMS Open Payments, the FDA Orange/Purple Books, and CDC WONDER each have their own API, format, and rate limits — and none of them share a schema. Pulling them together for an agent means solving each one and re-solving as they change. This server does that work once and exposes the result as agent-native tools.

The same NexGenData infrastructure that powers our standalone FDA, pharma, and healthcare actors — used by pharma competitive-intelligence teams, biotech analysts, med-device firms, and life-sciences investors — now sits behind a single streamable-HTTP endpoint.

The fifteen tools

Tool	What it does
search_drug_labels	FDA drug labels, approvals, safety (openFDA)
fda_drug_approvals	New FDA drug approvals (NDA/BLA/ANDA)
fda_adverse_events	FAERS drug + MAUDE device adverse-event reports
fda_recalls	FDA drug & device recalls and safety alerts
fda_warning_letters	FDA warning letters + Form 483 inspection findings
track_drug_shortages	Current and resolved FDA drug shortages
search_orange_book	Drug patents & exclusivity (generic-entry timing)
search_purple_book	Biologics & biosimilars + reference-product linkage
search_device_510k	FDA 510(k) medical-device clearances
track_pharma_catalysts	Pipeline catalysts and PDUFA decision dates
search_clinical_trials	ClinicalTrials.gov by condition, sponsor, phase
search_pubmed	PubMed biomedical literature
nih_grants	NIH RePORTER funded research grants
cms_open_payments	CMS Open Payments (Sunshine Act) pharma-to-physician
cdc_surveillance	CDC WONDER + FluView public-health surveillance

Every tool accepts a criteria object and a limit, and returns a uniform {tool, count, total_available, results} envelope — one parser for all fifteen.

Quick start

{
  "mcpServers": {
    "healthcare-fda": {
      "url": "https://nexgendata--healthcare-fda-intelligence-mcp.apify.actor/mcp",
      "headers": { "Authorization": "Bearer YOUR_APIFY_TOKEN" }
    }
  }
}

Call a tool:

{
  "name": "search_orange_book",
  "arguments": { "criteria": { "ingredient": "apixaban" }, "limit": 25 }
}

Common agent workflows

Generic-entry / patent-cliff analysis. Chain search_orange_book and search_purple_book to map when a drug's patents and exclusivity expire and which biosimilars reference it — the timing that drives generic competition.

Drug-safety monitoring. Run fda_adverse_events, fda_recalls, and fda_warning_letters on a product or manufacturer to build a live safety-signal feed.

Pharma catalyst calendar. Combine track_pharma_catalysts, fda_drug_approvals, and search_clinical_trials to track the events that move biotech valuations.

Payer & provider intelligence. Use cms_open_payments to surface pharma/device financial relationships with named physicians.

Research landscaping. Pair search_pubmed and nih_grants to map who is publishing and who is funded in a therapeutic area.

Pricing

Pay-per-event: $0.10 per tool call, $0.05 per bulk result (>10 records), negligible session-start fee. No subscription, no seat license. An agent firing 200 healthcare queries a month costs about $20 — versus the five- and six-figure annual contracts of IQVIA, Definitive Healthcare, or Citeline.

How this compares

vs. IQVIA / Citeline / Definitive Healthcare: built for enterprise seats and priced accordingly. This is built for agents, priced per call, and returns machine-ready JSON with source identifiers.

vs. raw openFDA + ClinicalTrials.gov + NIH + CMS: each is free but has a different API, format, and rate limit. This unifies fifteen of them behind one schema and one connection.

Companion servers

• SEC Corporate Events MCP — market-moving filings for the pharma/biotech names you're tracking.
• Legal & IP Intelligence MCP — litigation, FDA-adjacent enforcement, and IP.
• Environmental & ESG MCP — EPA, public-health, and nonprofit data.

FAQ

Do I need an FDA or NIH API key? No — only an Apify token for the MCP endpoint; upstream access is handled server-side.

How fresh is the data? Each tool queries its source live at call time.

Can I filter by date, drug, or sponsor? Yes — pass those in the criteria object; supported filters depend on the source.

Is this medical advice? No. This returns public regulatory and research data for informational use only — not medical advice.

Is the schema stable? Yes — every tool returns the same envelope; we version any changes.

Healthcare & FDA Intelligence MCP — by NexGenData. The FDA, NIH, CMS, and CDC, agent-ready.