COVID, Flu and RSV Wastewater Levels MCP Server
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
COVID-19, flu (influenza A) and RSV activity in the United States, read at call time from CDC open data on data.cdc.gov: National Wastewater Surveillance System viral activity levels for each sewershed site (Very Low to Very High), emergency department visits for acute respiratory illness by state, and the national percent of tests positive. CDC updates them weekly, usually on Fridays, for the week that ended the Saturday before. A model cannot know this week's levels. Use covid_flu_rsv_levels_state for "is COVID going around in Ohio" or "flu levels in Texas", covid_flu_rsv_levels_county for a county or city area, covid_flu_rsv_levels_national for the US picture, and covid_flu_rsv_compare_states before travel. Wastewater levels are not case counts. Not medical advice.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| covid_flu_rsv_levels_stateA | CDC wastewater viral activity levels for COVID-19, flu (influenza A) and RSV in one US state, DC or territory for the latest week: how many sewershed sites are Very Low, Low, Moderate, High or Very High, the share of people served by High or Very High sites, the median site level and the change from the week before, a 6-week trend, and the state's emergency department acute respiratory illness level. |
| covid_flu_rsv_levels_countyA | Wastewater levels for COVID-19, flu and RSV at every CDC sewershed site that serves a county: each site's level and category for the latest week, the readings for the last weeks (default 6, up to 12) and the population served. If no site serves the county, lists the counties in the state that have one. |
| covid_flu_rsv_levels_nationalA | The United States picture this week: the percent of laboratory tests positive for COVID-19, influenza and RSV with the weekly trend (default 8 weeks, up to 52), the share of people in sewersheds rated High or Very High for each virus, and emergency department acute respiratory illness activity with the highest states. |
| covid_flu_rsv_compare_statesA | Compare COVID-19, flu and RSV wastewater levels across 2 to 10 states for the latest week each state has reported, ranked by the share of people served by sites rated High or Very High, with each state's median site level and emergency department respiratory illness level. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 4 tools
Each tool is scoped by a distinct geographic unit: national, single state, single county, and a multi-state comparison. The descriptions make the boundaries explicit (aggregate national picture vs. per-state breakdown vs. per-site county detail vs. ranked cross-state comparison), so an agent can pick the right one without guessing.
All four names share the covid_flu_rsv_ prefix and end with a geographic scope modifier (national, state, county, states). The lone deviation is compare_states, which inserts a verb where the others use a bare scope noun, but it remains readable and predictable.
Four tools is well-matched to a focused public-health dataset: one per geographic aggregation needed plus a comparison view, with no redundant endpoints. Each tool covers a workflow that the others cannot.
The surface covers national, state, county, and cross-state views of the same metrics, which is the core lifecycle for this data. Minor gaps exist: county history is capped at 12 weeks while national allows 52, and there is no metadata/data-dictionary or arbitrary-time-window query tool, but agents can work around both.