Drug & Product Safety MCP Server
Pricing
from $15.00 / 1,000 tool calls
Drug & Product Safety MCP Server
Official FDA recalls, drug labelling, adverse event reports and ClinicalTrials.gov studies as MCP tools for AI agents. Every answer cites its source.
Pricing
from $15.00 / 1,000 tool calls
Rating
0.0
(0)
Developer
Tributary Data
Maintained by CommunityActor stats
0
Bookmarked
1
Total users
0
Monthly active users
5 hours ago
Last modified
Categories
Share
Give your AI agent direct access to FDA recall, labelling and adverse-event data plus ClinicalTrials.gov, so it can answer safety questions from official records rather than memory.
Every response includes a source pointing at the official dataset.
What it answers
- "Has this drug been recalled, and how serious was it?"
- "What does the official FDA label say about warnings for this medication?"
- "What adverse reactions are most reported for this drug?"
- "Has this brand of food been subject to a recall?"
- "What clinical trials are recruiting for this condition?"
Tools
| Tool | What it does |
|---|---|
search_drug_recalls | FDA drug recalls, with severity class and what that class means |
search_food_recalls | FDA food recalls by product, brand or firm |
search_device_recalls | FDA medical device recalls |
get_drug_label | The official FDA label: indications, dosage, warnings, boxed warnings, contraindications |
adverse_event_summary | Most frequently reported adverse reactions, across 20M+ reports |
find_clinical_trials | ClinicalTrials.gov studies by condition, intervention and recruiting status |
Built to be read correctly
Safety data is easy to misreport, so this server ships context alongside the numbers:
- Recall severity is explained, not just labelled. A "Class I" recall returns the FDA's actual meaning — reasonable probability of serious harm or death — rather than a code the agent has to guess at.
- Adverse event counts carry a caveat in the response. These are voluntary reports. They do not establish that a drug caused a reaction, and volume is heavily influenced by how widely a drug is prescribed. That warning travels with the data so it cannot be quietly dropped.
Why the data is trustworthy
All FDA data comes from openFDA, the agency's own public API. Trial data comes from ClinicalTrials.gov, run by the US National Library of Medicine. No third-party aggregators.
Pricing
$0.004 per tool call. No subscription, no minimum.
Limits and honesty
- US FDA data only. A product not recalled in the US may still be recalled elsewhere.
- openFDA data is not real-time; there is a lag between an action and its appearance in the API.
- Labels are the manufacturer's submitted content. Multiple labels may exist for one generic drug; this returns the closest match.
- This is not medical advice. It reports what regulators published. Clinical decisions need a clinician.
