Complete, runnable AIP implementations covering different industries and use cases. Each includes a manifest, server, and README.
Every example is a standalone Node.js server you can run locally. No dependencies beyond Node itself.
Metabolic health intake for men 40+. Anonymized health markers in, personalized plan recommendations out.
Binding health/assessment
Service marketplace that matches users with vetted local providers based on service type, location, and budget.
Binding service/matching
Pre-configured trial workspace setup based on team size, use case, and integration needs.
Binding saas/onboarding
Stateless health calculator. No binding, no account — submit numbers, get results.
Non-binding tool/calculator
Clone the repo and start any example in seconds.
git clone https://github.com/agent-intake-protocol/agent-intake-protocol.git
cd agent-intake-protocol/examples/health-assessment
node server.jsEach server exposes two endpoints:
GET /.well-known/agent-intake.json — Discovery manifestPOST /api/aip/<endpoint> — Intake submissionTest with curl:
# Discover
curl http://localhost:3000/.well-known/agent-intake.json
# Submit intake
curl -X POST http://localhost:3000/api/aip/metabolic-assessment \
-H "Content-Type: application/json" \
-d '{
"aip_version": "0.1.0",
"session_id": "test-001",
"intake_id": "metabolic-assessment",
"intake_data": {
"age_range": "40-49",
"sex": "male",
"primary_concern": "insulin_resistance",
"fasting_glucose_range": "elevated",
"activity_level": "sedentary",
"bmi_range": "overweight"
}
}'A metabolic health platform for men 40+. The agent submits anonymized health markers (age range, glucose range, BMI category) and receives a personalized optimization plan.
Key design choices: all inputs use ranges/categories instead of exact values, no PII required, and the privacy block explicitly declares data_retention: "none".
{
"aip_version": "0.1.0",
"provider": {
"name": "Man vs Health",
"url": "https://manvshealth.com",
"description": "Men's metabolic health optimization platform"
},
"intakes": [{
"id": "metabolic-assessment",
"name": "Metabolic Health Assessment",
"description": "Submit basic metabolic markers for a personalized plan",
"endpoint": "https://manvshealth.com/api/aip/metabolic-assessment",
"method": "POST",
"category": "health/assessment",
"input_schema": {
"type": "object",
"required": ["age_range", "sex", "primary_concern"],
"properties": {
"age_range": { "type": "string", "enum": ["30-39", "40-49", "50-59", "60+"] },
"sex": { "type": "string", "enum": ["male", "female"] },
"primary_concern": { "type": "string", "enum": ["weight", "energy", "insulin_resistance", "general"] },
"fasting_glucose_range": { "type": "string", "enum": ["normal", "elevated", "high", "unknown"] },
"activity_level": { "type": "string", "enum": ["sedentary", "light", "moderate", "active"] },
"bmi_range": { "type": "string", "enum": ["normal", "overweight", "obese", "unknown"] }
}
},
"offer_type": "personalized_recommendation",
"binding_available": true,
"requires_auth": false,
"privacy": { "data_retention": "none", "pii_required": false, "redacted_acceptable": true }
}]
}The server analyzes the intake data and returns a structured offer with a personalized plan, risk assessment, and recommended actions. If the user wants to proceed, they bind with their email and name.
{
"aip_version": "0.1.0",
"session_id": "test-001",
"status": "offer",
"offer": {
"id": "offer-abc123",
"summary": "Based on your profile, we recommend our Metabolic Reset program.",
"details": {
"plan": "Metabolic Reset — 12 Week Program",
"focus_areas": ["insulin sensitivity", "body composition", "energy optimization"],
"risk_level": "moderate",
"estimated_duration": "12 weeks",
"price": 197,
"currency": "USD"
},
"expires": "2026-03-06T12:00:00Z",
"bind_endpoint": "https://manvshealth.com/api/aip/bind",
"bind_requires": ["email", "full_name"]
}
}A marketplace that connects users with vetted local service providers. The agent submits a service request with type, zip code, and budget range — and gets back matched providers ranked by fit.
service/matchingsession (cleared after matching)Returns 1-3 matched providers with ratings, availability, and estimated price ranges. The user reviews matches before binding to share contact info with their chosen provider.
{
"status": "offer",
"offer": {
"summary": "We found 3 plumbers in your area.",
"details": {
"matches": [
{
"provider": "Johnson Plumbing",
"rating": 4.8,
"reviews": 142,
"available": "this_week",
"estimate": "$150–$300"
}
],
"total_matches": 3
},
"bind_requires": ["full_name", "phone", "address"]
}
}A project management tool that pre-configures trial workspaces based on team profile. The agent submits team size, use case, current tools, and needed integrations — and gets back a ready-to-go trial offer.
saas/onboarding30_days (trial duration)Returns a pre-configured workspace offer with selected plan, enabled integrations, migration path from current tools, and a direct setup URL at bind time.
{
"status": "offer",
"offer": {
"summary": "Your 14-day trial workspace is ready to configure.",
"details": {
"plan": "Team",
"seats": 15,
"features_enabled": ["kanban", "gantt", "time_tracking", "automations"],
"integrations": ["slack", "github"],
"migration_from": "jira",
"trial_days": 14
},
"bind_requires": ["email", "company_name"]
}
}The simplest possible AIP implementation — a stateless calculator. No binding, no account, no PII. Submit height and weight, get a BMI result. Demonstrates that AIP handles non-binding, utility-style interactions too.
tool/calculatorcomputed_resultfalse — no follow-up, no accountnone{
"status": "complete",
"offer": {
"summary": "Your BMI is 27.5 (Overweight).",
"details": {
"bmi": 27.5,
"category": "Overweight",
"healthy_range": "18.5–24.9",
"note": "BMI is a screening tool, not a diagnostic measure."
},
"expires": null,
"bind_endpoint": null
}
}These examples are starting points. AIP works for any business with an intake process — legal, financial, education, real estate, you name it.