CDC Disease Sentinel

CDC Disease Sentinel MCP Connector for Claude

A+

Detect disease surges in US public health data. CDC's weekly NNDSS case counts with trend, 52-week baseline, YTD comparison and a surge multiplier per disease, state or region — keyless, stateless.

5 tools Official Updated Oct 1, 2026 Official Vinkius Partner

CDC Disease Sentinel watches the CDC's National Notifiable Diseases Surveillance System so your agent can answer the question people keep asking: is anything spiking right now, and where?

What you can do

  • See what's surging — one call gives the latest week's case count, its average over a lookback window, and the multiplier: surge, elevated, or normal
  • Compare against the year — the dataset's own 52-week maximum and YTD vs the same period last year, with a plain-language verdict
  • Zoom into one state — national, any state, or a census region, with the exact area names returned so follow-ups never guess
  • Read the curve — the last N weeks of a disease as a compact series, so you see momentum, not just the latest number

How it works

  1. Subscribe to this server (no credential, no account — the dataset is public CDC open data)
  2. Ask "is anything spiking this week?" — the surge check runs against the configured disease(s)
  3. Re-run any time; the server computes everything from the live weekly series and keeps no state

Who is this for?

  • Public-health watchers — journalists, analysts, and the just-anxious curious who want to know what's actually moving, not headlines
  • Agents & automations — a scheduled "disease pulse" digest with deltas computed server-side
  • Researchers — reproducible weekly series with the source's own flags and caveats preserved
cdcdiseasesoutbreakspublic-healthnndsssurveillancesentinelhealth-data

5 tools expose this connector's capabilities to your AI agent.

list_diseases

The labels here are the exact strings the other tools expect — copy one into disease_trend, disease_baseline, surge_check or state_scan instead of inventing a name. MMWR years run July-to-June; use year to look at the previous MMWR year. The NNDSS diseases currently tracked by CDC weekly surveillance, with this week's national case counts

disease_baseline

Use this for the "is this unusual?" question — the ratio of YTD to the same period last year is the clearest signal. Where a disease stands right now: latest week vs the 52-week maximum and YTD vs the same period last year

disease_trend

Use this to see the shape of the curve: is the count climbing, flat, or collapsing? Weeks without a reported case show "no data" rather than a fabricated zero. Weekly case counts for one disease over the last N MMWR weeks, national or one state

state_scan

Areas with no reported cases show 0. The area names returned here are the exact strings to use as the state parameter of the other tools. Use this for "where is it concentrated?" questions, and to discover the exact area spelling for a follow-up. Where a disease is most reported in one week: every reporting area ranked by case count

surge_check

Verdicts: "surge" at or above threshold (default 2.0x), "elevated" at half the threshold, "normal" below that, plus special cases (no cases this week, or new cases where the baseline had none). This is the tool for "is X spiking right now?" — set threshold lower for early-warning sensitivity (e.g. 1.5) or higher to only flag true surges (3-5). Surge detector: latest week's count vs the average of the prior N weeks, with a multiplier verdict

See how to talk to your AI agent using CDC Disease Sentinel.

Is there a disease surge this week anywhere in the US?

Runs surge_check on the configured disease (or walks a few from list_diseases) and reports the multiplier verdicts.

Show me the last 16 weeks of national meningitis cases and say if it's trending up.

Uses disease_trend with weeks=16 and interprets the series shape.

Which states report the most norovirus this week?

Uses state_scan for norovirus and lists the ranked areas with case counts.

Compare this year's YTD tetanus against last year at the same point.

Uses disease_baseline and reports the YTD ratio with the plain-language verdict.

No. The dataset is public CDC open data on Socrata, readable without a token (anonymous access is throttled and capped at 500 rows per query — every tool stays well under that). No signup, no key, no per-user state.

Related Connectors