🏥 pack: clinical-pack

Ship clinical AI before FDA SaMD enforcement ships you.

Q: What human gates does clinical-pack introduce?

gate:samd-class (regulatory + clinical lead), gate:clinical-validation (clinical lead), gate:ide-approval (regulatory + sponsor). These layer on top of the standard plan/ship gates.

Q: What if my project doesn't match these signals exactly?

You can manually add the pack name to PROJECT.md or run /migrate to re-run detection with updated rules.

Building clinical decision support, an AI scribe, a diagnostic, or anything touching PHI? GreatCTO auto-attaches clinical-pack with FDA GMLP-10 + PCCP + SaMD classification (510(k) / De Novo / PMA) + EU AI Act Annex III medical + citation-grounding + subgroup fairness gates.

$ npx great-cto init All packs ↗

Auto-attach signals

Detected by CLI when:

fhir · hl7 · EHR/EMR · PHI · SaMD · CDS · scribe · radiology · pathology · telehealth-AI

The pack rides on top of your base archetype (web-service, ai-system, fintech, …) — it doesn't replace it. Auto-injects reviewer agents into the pipeline + opens human gates listed below.

Reviewer agents activated

2 specialists added to the pipeline.

01 · ai-clinical-reviewer

FDA GMLP-10 + PCCP + EU AI Act high-risk + hallucination + subgroup fairness + adversarial symptom

02 · fda-reviewer

IMDRF SaMD Class I/II/III/IV + 510(k)/De Novo/PMA path + IEC 62304 + ISO 14971

Human gates introduced

3 new gate types on top of `gate:plan` + `gate:ship`.

Gate	Owner	Trigger
`gate:samd-class`	regulatory + clinical lead	after fda-reviewer classifies
`gate:clinical-validation`	clinical lead	validation plan before patient data
`gate:ide-approval`	regulatory + sponsor	only for PMA path

Required artefacts before senior-dev claims tasks

8 concrete deliverables.

✓ Intended-use statement (autonomous vs assistive) in model card
✓ Predicate analysis document (or De Novo rationale)
✓ IEC 62304 software safety class declaration (A/B/C)
✓ ISO 14971 initial risk file
✓ Citation-grounding layer (PubMed/drug-label/guideline IDs ≥ 95% resolvable)
✓ Refuse-to-diagnose guardrail + clinician-in-loop trigger
✓ Subgroup fairness audit (sex × age × race) max-min AUC ≥ 0.8
✓ Drift-monitoring dashboard with weekly canary

EVAL suite required

5 golden-set scenarios shipped as templates.

Each EVAL has ≥5 test cases, pass threshold, regression interpretation, cross-refs to TM + gates. Run via your existing test framework.

EVAL-clinical-hallucination.md
EVAL-clinical-citation-grounding.md
EVAL-clinical-refuse-to-diagnose.md
EVAL-clinical-subgroup-fairness.md
EVAL-clinical-adversarial-symptom.md

Regulatory surface covered

8 standards / regulations addressed.

FDA GMLP (2021) FDA PCCP final guidance (2024) EU AI Act Annex III medical IEC 62304 ISO 14971 IEC 81001-5-1 EU MDR 2017/745 EU IVDR 2017/746

Real-world examples

30 companies in this space.

Intuitive Surgical

Robotic-assisted minimally invasive surgery (da Vinci)

publicUS

Ping An Good Doctor

Online healthcare for China

publicCN

Tempus AI

Precision medicine via AI + clinical data

publicUS

Vicarious Surgical

Surgical robots + VR controls

publicUS

Flatiron Health

Oncology EHR + real-world evidence

subsidiaryUS

Babylon Health

AI-powered telehealth

growthGB

Doctolib

European medical-appointment platform

growthFR

Halodoc

Indonesian telehealth + pharmacy

growthID

Iodine Software

AI for clinical revenue cycle

India's largest doctor-discovery + telehealth

growthIN

K Health

AI-powered primary care

series-eUS

Abridge

Generative AI for clinical conversations

series-cUS

Ada Health

AI symptom assessment

series-cDE

Diligent Robotics

Hospital service robots

series-cUS

Nabla

AI copilot for clinicians

series-cFR

Suki AI

Voice-enabled assistant for clinicians

series-cUS

Ambience Healthcare

AI scribe for primary care

series-bUS

Hippocratic AI

Safety-focused LLMs for healthcare

series-bUS

DeepScribe

Ambient AI medical scribe

series-aUS

OpenEvidence

Medical search engine for clinicians

series-aUS

Andromeda Surgical

Autonomous surgical robots

seedUS

AOA Dx

Blood test for early ovarian cancer detection

seedUS

Assort Health

Voice AI for patient + staff coordination

seedUS

careCycle

Voice AI for Medicare and ACA distributors

seedUS

Empirical Health

AI-scaled proactive primary care

seedUS

FutureClinic

Digital medical clinics that are 99% AI, 1% doctor

seedUS

Glass Health

AI for clinical decisions

seedUS

Knowtex

AI note-taking + charge capture for clinicians

seedUS

mdhub

AI assistants for mental-health clinic operations

seedUS

Listed companies operate in this space. Inclusion is based on publicly available product descriptions and does not imply endorsement of or by GreatCTO.

FAQ

Common questions about clinical-pack.

When does clinical-pack auto-attach?

When the CLI detects these signals in your repo: fhir · hl7 · EHR/EMR · PHI · SaMD · CDS · scribe · radiology · pathology · telehealth-AI. Override anytime by editing packs: in PROJECT.md.

What human gates does clinical-pack introduce?

gate:samd-class (regulatory + clinical lead), gate:clinical-validation (clinical lead), gate:ide-approval (regulatory + sponsor). These layer on top of the standard plan/ship gates.

What if my project doesn't match these signals exactly?

You can manually add the pack name to PROJECT.md or run /migrate to re-run detection with updated rules.

30 seconds

Drop GreatCTO into any repo — clinical-pack attaches automatically.

$ npx great-cto init

no signup·runs locally·pay your own API

Ship clinical AI before FDA SaMD enforcement ships you.

Detected by CLI when:

2 specialists added to the pipeline.

3 new gate types on top of gate:plan + gate:ship.

8 concrete deliverables.

5 golden-set scenarios shipped as templates.

8 standards / regulations addressed.

30 companies in this space.

Common questions about clinical-pack.

Drop GreatCTO into any repo — clinical-pack attaches automatically.

3 new gate types on top of `gate:plan` + `gate:ship`.