FDA Pharmaceutical Drug Data — MCP Server

Access US FDA pharmaceutical drug intelligence directly from AI agents. Search FDA drug labels, adverse event reports (FAERS), NDC codes, drug interaction descriptions, and FDA drug recall data through a native MCP interface. No API key required.

This MCP server connects AI agents to FDA openFDA APIs, enabling drug research, pharmacovigilance workflows, pharmaceutical due diligence, and clinical decision support without manual FDA API integration.

Features

• Drug Label Search — Query FDA drug label database for indications, contraindications, dosage, warnings, black box warnings, and active ingredients
• Adverse Event Reporting — Search FDA FAERS (FDA Adverse Event Reporting System) for drug safety signals, serious outcomes, and pharmacovigilance data
• NDC Code Lookup — Find National Drug Codes and associated product information including strength, package size, and manufacturer
• Drug Interaction Checking — Retrieve drug-drug interaction descriptions from FDA data for clinical decision support
• Drug Recall Tracking — Monitor FDA drug recalls by manufacturer, product type, or drug class (Class I, II, III)
• Native MCP Interface — Standard Model Context Protocol integration for AI agents (Claude Desktop, Cursor, Windsurf, other MCP clients)
• No API Key Required — FDA openFDA data is public; this actor uses Apify PPE for billing without separate FDA credentials
• Pay Per Use Pricing — No subscription; only pay for the API calls you actually make

Quick Start

Add to your MCP client configuration:

{
  "mcpServers": {
    "drug-intelligence-mcp": {
      "url": "https://drug-intelligence-mcp.apify.actor/mcp"
    }
  }
}

Once configured, AI agents can call drug intelligence tools directly in conversation.

Tools

Tool	Price	Description
search_drug_labels	$0.03	Search FDA drug label database
get_drug_adverse_events	$0.08	Search FDA adverse event reports (FAERS)
search_drug_by_ndc	$0.03	NDC code lookup
get_drug_interactions	$0.05	Drug-drug interaction descriptions
search_drug_recalls	$0.03	Track FDA drug recalls

search_drug_labels

When to call: AI agent researching drug indications, contraindications, or active ingredients for clinical or investment research.

Example AI prompt: "Show me the FDA label for Keytruda (pembrolizumab) — what are its indications and black box warnings?"

Input parameters: drug_name (required), max_results (optional, default 5)

get_drug_adverse_events

When to call: AI agent investigating drug safety signals, adverse event patterns, or pharmacovigilance data.

Example AI prompt: "What are the most serious adverse events reported for Ozempic (semaglutide) in the last 2 years?"

Input parameters: drug_name (required), seriousness (optional: "death", "serious", "other"), max_results (optional, default 10)

search_drug_by_ndc

When to call: AI agent verifying NDC codes for formulary management, insurance verification, or pharmaceutical logistics.

Example AI prompt: "Look up NDC code 0001-1234-01 — what's the drug name, strength, and package size?"

Input parameters: ndc_code (required)

get_drug_interactions

When to call: AI agent checking potential drug-drug interactions before clinical decision support or formulary review.

Example AI prompt: "What interactions exist between warfarin and aspirin according to FDA data?"

Input parameters: drug_name (required)

search_drug_recalls

When to call: AI agent tracking drug recall status for pharmacy risk management or supply chain intelligence.

Example AI prompt: "Show me all FDA drug recalls involving blood pressure medications in the last 12 months."

Input parameters: drug_name (optional), firm (optional), product_type (optional), max_results (optional, default 20)

Input Parameters

Parameter	Type	Required	Description
drug_name	string	Varies by tool	Drug brand or generic name
ndc_code	string	For NDC tool	National Drug Code (format: XXXX-XXXX-XX)
seriousness	string	No	Filter for adverse events: "death", "serious", "other"
firm	string	No	Manufacturer name for recall search
product_type	string	No	Product type for recall search
max_results	integer	No	Maximum results to return (default varies by tool)

Output

Each tool returns structured JSON with the relevant FDA data, source attribution, and pagination metadata:

{
  "query": "Keytruda",
  "total_results": 3,
  "labels": [
    {
      "brand_name": "KEYTRUDA",
      "generic_name": "pembrolizumab",
      "indications": "melanoma, NSCLC, renal cell carcinoma",
      "warnings": ["Grade 4 or life-threatening pneumonitis", "severe colitis"],
      "active_ingredients": ["pembrolizumab 100mg"],
      "dosage_and_administration": "200mg IV every 3 weeks",
      "source": "FDA Drug Label"
    }
  ]
}

Pricing

Tool	Per Call
search_drug_labels	$0.03
get_drug_adverse_events	$0.08
search_drug_by_ndc	$0.03
get_drug_interactions	$0.05
search_drug_recalls	$0.03

No subscription. Pay per use via Apify Pay-Per-Event (PPE). No API keys required — FDA openFDA data is public domain.

Data Sources

Source	Coverage	What's Available
FDA openFDA Drug Labels	US	Indications, dosage, warnings, ingredients
FDA FAERS	US	Adverse event reports, serious outcomes
FDA NDC Directory	US	National Drug Codes, package sizes
FDA Drug Recalls	US	Class I/II/III pharmaceutical recalls

Troubleshooting

"No results found" for a common drug name: FDA drug labels use standardized names. Try the generic name instead of the brand name (e.g., "pembrolizumab" instead of "Keytruda"). Also try common misspellings and alternative spellings.

"NDC code not found" error: NDC codes follow format XXXX-XXXX-XX (11 characters with hyphens). Some older products use 10-digit format without leading zeros. Check the FDA NDC directory for the correct code format.

High pricing on adverse event queries: FAERS queries are $0.08 per call because FDA returns large datasets. Use seriousness: "death" or max_results: 5 to limit response size and cost.

MCP client not connecting: Verify the MCP server URL is exactly https://drug-intelligence-mcp.apify.actor/mcp with no trailing slash. Some MCP clients require a restart after configuration changes.

Timeout errors on large queries: Drug label and recall searches may timeout on very broad queries. Narrow your search by adding a specific drug name, manufacturer, or date range.

How It Compares to Alternatives

Aspect	Drug Intelligence MCP	FDA openFDA API	RxNorm API
Price	$0.03–$0.08/call	Free	Free
AI agent integration	Native MCP	REST only	REST only
Adverse events	Yes (FAERS)	Yes (raw data)	No
Drug interactions	Yes	Yes	No
Recall tracking	Yes	Yes	No
No API key required	Yes	Yes	Yes

Connect to AI Agents

Claude Desktop

Add to your Claude Desktop MCP settings file (typically at ~/Library/Application Support/Claude/claude_desktop_config.json):

{
  "mcpServers": {
    "drug-intelligence-mcp": {
      "url": "https://drug-intelligence-mcp.apify.actor/mcp"
    }
  }
}

Cursor

Add to Cursor MCP settings in Settings → AI → MCP Servers:

Name: drug-intelligence-mcp
URL: https://drug-intelligence-mcp.apify.actor/mcp

Windsurf

Add via Windsurf Settings → MCP → Add new MCP server with the same URL.

SEO Keywords

FDA drug label API, adverse event reporting system, FAERS, NDC code lookup, drug recall database, FDA openFDA, AI agent pharmaceutical research, drug safety signals, pharmacovigilance API, drug interaction checker, clinical decision support, pharmaceutical intelligence, FDA drug database, drug recall tracker, MCP server for AI agents