RCM & insurance

Validate a candidate, investigate refusal and read a non-posting simulation

Use the exact server qualification and explicit supplied-input simulation without replacing missing evidence with pass flags.

Current source · Source reviewed; exact stated local synthetic UI/API continuation and independent readback observed · 8 min read

Before you begin

Use a separate synthetic training estate and authorized maker/reviewer accounts for demonstrations. For production, use the actual approved organizational and contract sources.

1. Validate the exact publication candidate

Open Rule Publishing → Packages → Publish package. Name the candidate and run Validate configuration. The server checks model constraints, approvals, dates, evidence grade, retained-byte hashes and active gate invariants against the exact saved manifest.

**Run server checks on the exact candidate · 2026-10-18**Actual audited MANSOQ production frontend in isolated synthetic training. The maker opens Publish package, supplies a unique label and runs Validate configuration. No approval code was entered in this screenshot. A client pass flag or a pasted test reference is not accepted.
**Server candidate validation**: The normal endpoint binds checks to the exact candidate manifest.

2. Read and preserve all qualification failures

The protected Polyclinic example checked 25 reviewed settings, all refused because legacy synthetic URI evidence is unreadable. Its 25 unresolved gates remain recorded. This failure is valid teaching evidence; a stored review flag, client pass flag or arbitrary test reference is not a replacement for actual retained files.

**Recognize retained-source validation failures · 2026-10-18**Actual audited MANSOQ production frontend in isolated synthetic training. Actual server qualification checked 25 existing Polyclinic settings and refused all because their legacy synthetic URI sources have no readable retained bytes. All 25 gates remain. At 1280x720 the 2870px dialog overflows and clips header/footer; this frame shows the middle of the refusal list, not the complete report.
**Readable source required**: A saved review flag alone cannot replace retained content.
**Each row retains a finding**: Keep the full authenticated qualification report for investigation.

3. Enter explicit simulation facts and read the scope

In Simulate, enter label and before/after currency, coverage decision, total, payer and patient amounts for each representative outpatient/pharmacy/inpatient case. The example intentionally compares supplied 150/120/30 SAR with 155/124/31 SAR. Read posted nothing and separate +1 patient/+4 payer changes. No resolver enumerated affected policies, services or encounters, so those stay not resolved rather than none.

**Read an explicit manual-allocation simulation · 2026-10-18**Actual audited MANSOQ production frontend in isolated synthetic training. The actual UI compares three manually entered fictional outpatient/pharmacy/inpatient allocation pairs: total 150→155, payer 120→124, patient 30→31 SAR. Server says posted nothing. This is supplied-input delta comparison, not a real tariff reprice, payer answer or automatic enumeration of affected policies/services/encounters.
**Posted nothing**: Simulation reports no charge, claim or balance posting.
**Patient and payer deltas**: Read the separate +1 and +4 supplied-input changes.
**Before and after**: These figures were explicitly typed; no default entitlement was inferred.

4. Use a genuinely isolated candidate for a separate successful demonstration

The separate Clinic teaching package contains one dated source-linked organization-master row, actual retained mock bytes, separate reviewer approval and a current server check. Its green result does not repair or qualify the protected Polyclinic legacy rows. Follow the separate guarded publication guide for that observed path.

**Confirm exact server qualification passed · 2026-10-18**Actual audited MANSOQ production frontend in isolated synthetic training. The maker UI server checks passed for exactly one reviewed organization-master setting using actual retained fictional bytes, approval and dates. The empty reviewer-code field remains required. Qualification lasts 15 minutes and must be rerun after any configuration/evidence change.
**Exactly one reviewed setting**: Success is scoped to this candidate.
**Expiry and change binding**: The receipt cannot validate a different or changed candidate.
**Independent one-use approval still needed**: The code is private and was not included in this screenshot.

**Verify actual guarded package publication · 2026-10-18**Actual audited MANSOQ production frontend in isolated synthetic training. Actual maker UI Publish saved immutable sequence 1 with a distinct existing Clinic approver and current server-issued qualification. The one-use reviewer code was issued normally and never included in screenshots. No gate was open in this separate one-row Clinic candidate; this says nothing about the protected Polyclinic gates.
**Immutable sequence 1**: A correction must become a new package.
**Different publisher and approver**: Normal dual control remained enforced.
**Server qualification reference**: The package binds the actual successful candidate check.
**Candidate-scoped gates**: No open gates in this Clinic candidate does not close another facility’s gates.

Source evidence

References are pinned to the private MANSOQ source repository. Repository access is required to open these links.