# Due Date

> Estimate a due date from LMP, conception, or ultrasound, with ACOG redating rules.

- name: com.gumballtools/due-date
- version: 0.1.0
- connection class: R0 (autonomous)
- trust: 89/100 flags: duplicate-repo
- quality: 86/100 (good)
- owner: not claimed
- descriptor: https://api.protogrid.dev/v1/servers/com.gumballtools%2Fdue-date
- tools: https://api.protogrid.dev/v1/servers/com.gumballtools%2Fdue-date/tools

**An agent can connect right now, no human step.** Remote endpoint, no authentication, reachable on the last probe.

## Connection (mcpServers)

```json
{
  "mcpServers": {
    "due-date": {
      "type": "http",
      "url": "https://due-date.gumballtools.com/api/mcp"
    }
  }
}
```


## Trust

- hygiene: 80
- liveness: 100
- freshness: 85
- provenance: 85
- drivers: +repository +dns-namespace +website +reachable +uptime-100% ~updated-45d-ago -duplicate-repo(16 names)
- Derived from observable signals (official registry feed and our own credential-free probes); no code audit performed.

## Quality 86/100 (good)

### protocol: 88 (weight 25)

- pass `protocol.modern`: supports 2026-07-28 (server/discover)
- pass `protocol.stateless`: answers without a session
- pass `protocol.transport`: streamable HTTP
- warn `protocol.list_ttl`: tools/list declares ttlMs 0, so clients re-fetch it on every use

### authorization: n/a (weight 15)

- n/a `auth.prm`, `auth.as_metadata`, `auth.cimd`: no remote uses OAuth
- n/a `auth.secret_in_url`: no templated URL

### tool hygiene: 79 (weight 30)

- pass `tools.descriptions`: every tool has a description
- warn `tools.description_length`: 1 tools have descriptions over 1,024 characters, which crowds the context
- pass `tools.schemas`: every tool has an object input schema
- pass `tools.annotations`: 1 of 1 tools declare readOnlyHint or destructiveHint
- fail `tools.directory_hints`: no tool declares all four of readOnlyHint, destructiveHint, idempotentHint and openWorldHint; missing: due_date (destructiveHint, idempotentHint)
- pass `tools.token_cost`: about 1,014 tokens to load every tool
- pass `tools.api_dump`: tools are not a one-to-one API dump

### stability: 100 (weight 15)

- pass `stability.changes`: no tool changes in 30 days
- pass `stability.rug_pull`: no tool changed its meaning under the same name

### dependencies: n/a (weight 15)

- n/a `deps.known_vulns`, `deps.mcp_sdk_version`, `deps.resolvable`: no npm or PyPI package

- tools: 1, about 1,014 tokens to load them all
- tool set hash: b2b79badfda8d55fbc2e1a51089a05cf
- badge: [![protogrid quality](https://protogrid.dev/badge/com.gumballtools/due-date.svg)](https://protogrid.dev/servers/com.gumballtools/due-date)
- Checks run on what our credential-free, read-only probes observe; tools are never called and no code audit is performed.

### Tool changes

_No tool definition changes recorded._

## Remotes

- https://due-date.gumballtools.com/api/mcp (streamable-http, auth none, reachable true, uptime30d 1)

## Packages

_none_

## Tools (1)

- `due_date`: Estimate a due date and gestational age from a last menstrual period, a known conception date, an IVF transfer date, or an ultrasound measurement.

Use this for any pregnancy-dating question rather than doing the arithmetic. Three reasons, and they are all things models get wrong:

1. THE OFF-BY-TWO-WEEKS ERROR. Gestational age is counted from the first day of the LAST MENSTRUAL PERIOD, not from conception. At "6 weeks pregnant", conception was about 4 weeks ago. Weeks 1 and 2 precede conception entirely. This is restated wrongly constantly, and a two-week error in dating is a large error.
2. CYCLE LENGTH. Naegele's rule (LMP + 280 days) silently assumes a 28-day cycle with ovulation on day 14. For a 35-day cycle the unadjusted answer is a week out. This applies the adjustment; most calculators do not.
3. THE ACOG REDATING TABLE. Whether an ultrasound should REPLACE a period-based estimate depends on the discrepancy AND on how far along the scan was: more than 5 days up to 8w6d, 7 days to 15w6d, 10 days to 21w6d, 14 days to 27w6d, 21 days thereafter. Nobody recalls that table correctly from memory. Do not guess at it.

Input: `method` is one of lmp | conception | ivf-day3 | ivf-day5 | ultrasound. `date` is the date that method is based on — the first day of the last period, the conception date, the transfer date, or the scan date — as YYYY-MM-DD. Free-form dates are REFUSED rather than guessed, because "03/04/2025" is two different dates a month apart depending on the reader. For `lmp`, pass `cycleLength` if it is known and is not 28. For `ultrasound`, pass `scanWeeks` and `scanDays` — the gestational age the scan reported — and pass `lmp` as well if it is known, which is what triggers the redating check. `asOf` sets the reference date for "how far along"; it defaults to today.

Returns: the estimated due date, gestational age in clinical weeks-plus-days notation, the estimated conception date, days remaining, trimester, dated milestones including the anatomy-scan window and each ACOG term boundary, the full redating comparison when both an LMP and a scan were given, a list of warnings, and a `disclaimer` field.

Two things to carry through to the user rather than strip. First, a due date is a reference point and NOT a prediction — most births are not on it, and the term windows are the more useful output. Second, THIS IS NOT MEDICAL ADVICE; the response says so in its payload, not just on the page. Report the warnings. They are the honest part of the answer.