Install any skill in seconds. Free to start, no credit card required.
Get Started Free →Given a newborn's five delivery-room findings described in prose — skin color, heart rate, reflex response to stimulation, muscle tone, and breathing effort — computes the total Apgar score (0-10) by scoring each of the five signs 0, 1, or 2 and summing them. Use when someone reports what a just-born infant looked like at the one- or five-minute check and wants the single vitality number. Do NOT use for estimating gestational age (Ballard/Dubowitz), respiratory distress scoring (Silverman-Andersen), adult vital-sign triage, or deciding resuscitation steps.
.claude/skills/apgar-score-compute/SKILL.md| Model | Eval pass | Runs |
|---|---|---|
| gemini-3.6-flash | 100% | 7 |
| Model | Lift | Δ tokens | Δ turns | Cases | Verified |
|---|---|---|---|---|---|
| gemini-3.5-flashbest | +24% | — | 0% | 25 | 61d ago |
| gemini-3.6-flash | — | +166% | 0% | 22 | 54d ago |
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-15 | ✗→✗ | = Same ✗ | — | — |
| case-09 | ✗→✗ | = Same ✗ | — | — |
| case-04 | ✗→✗ | = Same ✗ | — | — |
| case-07 | ✗→✗ | = Same ✗ | — | — |
| case-14 | ✗→✗ | = Same ✗ | — | — |
Handed a description of a newborn a minute after delivery, the base model tends to eyeball a number. Two failures dominate: it treats blue hands and feet on an otherwise pink baby as full color credit (it is not), and it rounds a heart rate in the 90s up to full credit (it is not). It also silently drops a sign it was not sure how to score, so a five-sign total comes back built from four. The score is not a gestalt impression — it is a sum of five fixed sub-scores, and each one has a hard rule.
The Apgar score rates a newborn's condition as an integer from 0 to 10. You score five signs, each 0, 1, or 2, and add them. The signs are Appearance (color), Pulse (heart rate), Grimace (reflex irritability), Activity (muscle tone), and Respiration (breathing effort). Source: Apgar V. (1953), A proposal for a new method of evaluation of the newborn infant, Curr Res Anesth Analg 32:260. This is the standard clinical rubric, US/international, as of 2026-07; it has been stable since introduction.
Score all five. A missing or unmentioned sign is scored 0, not skipped.
Appearance (skin color)
Pulse (heart rate)
Grimace (reflex irritability to stimulation)
Activity (muscle tone)
Respiration (breathing effort)
Total = Appearance + Pulse + Grimace + Activity + Respiration. Report the integer only.
Acrocyanosis ceiling. Trunk pink, hands and feet still dusky; rate 132; screams when the foot is flicked; limbs well flexed and thrashing; a loud cry.
Heart rate in the 90s. Pink all over; rate charted at 94; vigorous cry to suction; active motion; strong cry.
A dropped sign. Pale throughout; no heartbeat found; a weak grimace to the sole flick; some flexion of the legs; feeble irregular gasps.
One sign absent, the rest strong. Fully pink; rate 158; coughs and cries hard; active and well flexed; but not breathing at all.
Return the total as a JSON object keyed apgar with an integer value, e.g. {"apgar": 9}. No sub-scores, no prose, unless the caller asks for the breakdown.
scripts/grading_spec.json holds the correct integer for each benchmark newborn, and scripts/grade.py parses the answered apgar value and compares it exactly. See the eval suite.
| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-15 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-09 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-04 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-07 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-14 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-10 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-12 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-02 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-05 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-03 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-11 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-19 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-08 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-06 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-22 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-21 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-16 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-01 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-20 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-13 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-17 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
case-18 | fail→fail | — | — | — | — | — | — | — | — | — | — | — | — |
DecimalAI ran this skill against gemini-3.5-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted. The headline lift of +24 percentage points is the difference between those two pass rates over the 22 comparable cases.
The per-case answers from this run were removed by the retention sweep, so the case table below shows the verdicts without the text either arm produced. The counts above were recorded at the time and are unaffected. Answers are now kept for 180 days.
| Model | Method | Date | Lift |
|---|---|---|---|
| gemini-3.5-flash | replay | 7/21/2026 | +24% |
| gemini-3.5-flash | replay | 7/21/2026 | +24% |
Other measured skills in the registry, with their headline benchmark lift.