Insurance acceptance testing

Use one isolated Sandbox project and synthetic records for the full payer pilot. Keep the API key in a secret manager and store the request ID, idempotency key and resulting resource identifier for every write.

Run the pilot as one shared evidence room. Payer engineering, actuarial and finance, Care operations, security and privacy, and governance each sign off on their own controls against the same synthetic workflow.

Pilot control room

Engineering
Actuarial and finance
Care operations
Security and privacy
Governance

Own the contract commit, environment mapping, retry policy, idempotency keys, Webhook consumer and integration logs.

Acceptance sequence

Sequence	Test	Evidence to retain
1	Record a versioned Coverage observation	Source reference, source version, observation ID and snapshot version
2	Replay the same observation	Same response and `replayed` behavior, with no second observation
3	Reuse the key with different content	Deterministic conflict response
4	Check eligible and ineligible services	Decision, reason codes, effective time and limits
5	Approve a required pre-authorization twice	One authorization and one Benefit reservation
6	Submit a Claim for delivered Care	Claim ID, submission version, service code and evidence references
7	Request and submit bounded evidence	Information request, evidence references and new submission version
8	Reject an unbalanced adjudication	Validation error with no financial state change
9	Record a balanced partial approval	Policy version, line decisions, reasons and payer amount
10	Submit remittance advice	Allocation evidence without a settlement claim
11	Match an independent settlement observation through the approved pilot adapter	Reconciled remittance, balanced ledger evidence and settled Claim state
12	Deliver a duplicate Webhook	One business action despite repeated delivery

Authorization tests

Run each request with the expected scope, without the scope, against another Organization and against a production identifier from the Sandbox project. Expected results:

the correct scope and tenant can reach the resource;
a missing scope returns 403;
a foreign or unavailable resource returns 404 without confirming its existence;
Sandbox credentials cannot act on production data;
payer resources do not expose Person identity or clinical content.

Failure and retry tests

Exercise:

malformed JSON and unknown fields;
an invalid service code or source version;
expired Coverage and exhausted Benefits;
authorization for the wrong Booking or service;
Claim submission before Session delivery;
missing requested evidence;
duplicate Claim references;
currency mismatch and remittance over-allocation;
429 handling with backoff;
a timeout followed by a same-key retry;
Webhook signature failure, stale timestamp and duplicate event delivery.

Financial assertions

For every adjudicated Claim, independently recompute:

line.allowed = line.payer + line.patient_responsibility
line.billed = line.allowed + line.adjustment
claim totals = exact sums of line amounts
advice outstanding = approved payer amount - cumulative remittance allocations
settlement outstanding = approved payer amount - cumulative matched settlement

Reject the pilot if the API accepts an unbalanced amount, changes the result of a same-key replay, crosses an Organization boundary or represents remittance advice as settlement.

Operational evidence pack

The pilot evidence pack should contain:

the OpenAPI contract commit used by both teams;
the payer-to-service-code mapping and its version;
environment, Organization and scope mapping;
synthetic request and response fixtures;
authorization and negative-test results;
idempotency and retry results;
Webhook verification results;
financial reconciliation results;
named owners for payer operations, security, privacy and incident response;
the production activation and rollback checklist.

Do not place keys, personal data, clinical records or payment credentials in the evidence pack.

Minimum evidence pack structure

01-contract

openapi-commit.txt

service-code-map.csv

scope-map.md

02-execution

03-financial-controls

04-governance

Evidence that must stay outside the pack

API keys, WorkOS credentials, Person identity, Clinical Notes, Session content, Assessment answers, payment credentials and unrestricted production exports remain outside the pilot repository.

Exit criteria

The integration is ready for production approval when both teams can reproduce the full sequence, every negative test fails closed, financial totals reconcile, audit evidence is retrievable and the named owners approve activation. Request Sandbox access to begin the pilot.

Payer workflows

Integration assurance

Insurance acceptance testing

Pilot control room

Acceptance sequence

Authorization tests

Failure and retry tests

Financial assertions

Operational evidence pack

Exit criteria

​Pilot control room

​Acceptance sequence

​Authorization tests

​Failure and retry tests

​Financial assertions

​Operational evidence pack

​Exit criteria

Pilot control room

Acceptance sequence

Authorization tests

Failure and retry tests

Financial assertions

Operational evidence pack

Exit criteria