Files
cyclone/docs/superpowers/specs/2026-06-21-cyclone-claims-unique-constraint-and-409-ux-design.md
T

386 lines
14 KiB
Markdown

# Drop `UNIQUE(batch_id, patient_control_number)` on Claims + Robust 409 UX
**Date:** 2026-06-21
**Branch:** `main`
**Status:** Draft (brainstorming approved, awaiting writing-plans)
**Scope:** Backend schema + minimal frontend error handling for collision 409.
---
## 1. Why this exists
Today, every multi-claim 837P file in `docs/prodfiles/837p-from-axiscare/` and
`docs/prodfiles/FromHPE/` returns **HTTP 409** on upload. The cause:
* The `claims` table has `UNIQUE(batch_id, patient_control_number)` declared
inline in `CREATE TABLE` (auto-index `sqlite_autoindex_claims_2`).
* `store._claim_837_row` populates `patient_control_number` from
`claim.subscriber.member_id` (the subscriber's insurance member id),
not from the CLM01 segment value.
* A real 837P submission routinely contains many `CLM*` segments per
subscriber (a member seeing multiple providers on the same day). All
those rows share the same `member_id` → same `patient_control_number`.
Within one batch, the constraint fires on claim #2.
Migration `0003_drop_claims_remits_unique_constraints.sql` already exists
with the right intent but is **buggy**: it does
`DROP INDEX IF EXISTS uq_claims_batch_pcn`, but the constraint is inline,
not a named index — so the `DROP INDEX` is a no-op and the constraint
survives.
The 409 error message ("duplicate CLM01 control numbers") is also
misleading because the column stores `member_id`, not CLM01.
This SP fixes both: drops the constraint properly via SQLite table
recreation, and gives the upload page a structured error panel for the
collision case so the operator can find the existing batch in one click.
---
## 2. Operator surface
| Surface | Change |
|---|---|
| DB | New migration `0013_drop_claims_unique_constraint.sql`; idempotent. |
| Python | `store.find_existing_batch_for_claim(claim_id)` new helper. |
| API | `POST /api/parse-837` and `/api/parse-835` 409 responses gain `existing_batch_id`. |
| UI | `/upload` page renders an inline error panel for 409 with a link to the existing batch. |
| Tests | New migration test, store helper test, API test, component test. |
No CLI / settings changes.
---
## 3. Schema migration
`backend/src/cyclone/migrations/0013_drop_claims_unique_constraint.sql`:
```sql
-- version: 13
-- Drop the inline UNIQUE(batch_id, patient_control_number) on claims.
-- Migration 0003 attempted DROP INDEX IF EXISTS uq_claims_batch_pcn but
-- the constraint is inline in CREATE TABLE, so the drop was a no-op.
-- The only way to remove an inline UNIQUE in SQLite is table recreation.
--
-- X12 837P allows any number of CLM segments per 2000B subscriber loop;
-- claim identity is provided by the primary key (claims.id = CLM01).
-- The remittances table had a parallel constraint already removed in 0003
-- (because that one WAS a named index), so this migration only touches
-- claims.
--
-- The migration runner (db_migrate.py) wraps each .sql file in an
-- implicit transaction via engine.begin(), so we MUST NOT use BEGIN/COMMIT
-- inside the file (nested transactions fail in SQLite). We defer FK
-- enforcement with PRAGMA defer_foreign_keys instead of turning FKs off
-- (which is a no-op inside a transaction in SQLite). The deferred
-- checks fire at commit and validate against the renamed claims table.
PRAGMA defer_foreign_keys = ON;
CREATE TABLE claims_new (
id TEXT PRIMARY KEY,
batch_id TEXT NOT NULL REFERENCES batches(id) ON DELETE CASCADE,
patient_control_number TEXT NOT NULL,
service_date_from DATE,
service_date_to DATE,
charge_amount NUMERIC(12, 2) NOT NULL DEFAULT 0,
provider_npi TEXT,
payer_id TEXT,
state TEXT NOT NULL DEFAULT 'submitted',
state_before_reversal TEXT,
matched_remittance_id TEXT REFERENCES remittances(id),
raw_json TEXT,
rejection_reason TEXT,
rejected_at TIMESTAMP,
resubmit_count INTEGER NOT NULL DEFAULT 0,
state_changed_at TIMESTAMP,
payer_rejected_at TEXT,
payer_rejected_reason TEXT,
payer_rejected_status_code TEXT,
payer_rejected_by_277ca_id TEXT,
payer_rejected_acknowledged_at TEXT,
payer_rejected_acknowledged_actor TEXT
-- NO UNIQUE (batch_id, patient_control_number) — removed.
);
INSERT INTO claims_new SELECT * FROM claims;
DROP TABLE claims;
ALTER TABLE claims_new RENAME TO claims;
-- Recreate secondary indexes (same names, same columns as initial schema
-- plus later migrations).
CREATE INDEX ix_claims_state ON claims(state);
CREATE INDEX ix_claims_patient_control_number ON claims(patient_control_number);
CREATE INDEX ix_claims_service_date_from ON claims(service_date_from);
CREATE INDEX ix_claims_state_changed_at ON claims(state, state_changed_at);
CREATE INDEX idx_claims_payer_rejected_at ON claims(payer_rejected_at);
CREATE INDEX idx_claims_payer_rejected_unack
ON claims(payer_rejected_at)
WHERE payer_rejected_acknowledged_at IS NULL;
```
**Note on the `db.py` ORM model:** `backend/src/cyclone/db.py` declares
the `Claim` ORM model with `UNIQUE(batch_id, patient_control_number)` in
`__table_args__`. After this migration the DB no longer enforces that
constraint, so the ORM declaration becomes aspirational. We leave it in
place as documentation (and as a guard if the DB ever gets rebuilt from
ORM on a fresh schema). No code reads through it; the dedup happens via
PK on `claims.id` in `store.add`.
---
## 4. Store helper
New function in `backend/src/cyclone/store.py`:
```python
def find_existing_batch_for_claim(claim_id: str) -> str | None:
"""Return the batch_id of the first batch containing this CLM01, or None."""
from cyclone import db
from cyclone.db import Claim
with db.SessionLocal()() as s:
row = s.execute(
select(Claim.batch_id).where(Claim.id == claim_id).limit(1)
).first()
return row[0] if row else None
def find_existing_batch_for_remit(payer_claim_control_number: str) -> str | None:
"""Return the batch_id of the first batch containing this CLP01, or None."""
from cyclone import db
from cyclone.db import Remittance
with db.SessionLocal()() as s:
row = s.execute(
select(Remittance.batch_id)
.where(Remittance.id == payer_claim_control_number)
.limit(1)
).first()
return row[0] if row else None
```
Both pure reads; no transaction management needed.
---
## 5. API change
`backend/src/cyclone/api.py`:
### 837 path (line 394-415)
After the UNIQUE constraint is dropped in migration 0013, an `IntegrityError`
in `store.add` can only originate from the PK on `claims.id` (CLM01) —
either two claims in the same file share CLM01 (rare) or the same CLM01
exists in a prior batch. The handler picks the first claim from
`result.claims` and asks the helper whether a prior batch already holds
that CLM01:
```python
except IntegrityError as exc:
first_claim_id = result.claims[0].claim_id if result.claims else None
existing_batch_id = (
store.find_existing_batch_for_claim(first_claim_id)
if first_claim_id else None
)
body = {
"error": "Duplicate claim",
"detail": (
"This file (or one previously ingested with the same "
"claim control number) collides with an existing record. "
"Inspect the file for duplicate CLM01 control numbers, or "
"remove the existing batch before retrying."
),
"batch_id": rec.id,
}
if existing_batch_id and existing_batch_id != rec.id:
body["existing_batch_id"] = existing_batch_id
log.warning("Duplicate claim while persisting batch %s: %s", rec.id, exc)
return JSONResponse(status_code=409, content=body)
```
The `existing_batch_id != rec.id` guard avoids surfacing the just-failed
batch as a "previous" batch (it never persisted).
### 835 path (line 588-602)
Same pattern with `find_existing_batch_for_remit` and the first remit's
`payer_claim_control_number` (`result.claims[0].payer_claim_control_number`):
```python
except IntegrityError as exc:
first_pcn = result.claims[0].payer_claim_control_number if result.claims else None
existing_batch_id = (
store.find_existing_batch_for_remit(first_pcn)
if first_pcn else None
)
body = {
"error": "Duplicate remittance",
"detail": (
"This 835 file (or one previously ingested with the same "
"payer claim control number) collides with an existing record. "
"Remove the existing remittance before retrying."
),
"batch_id": rec.id,
}
if existing_batch_id and existing_batch_id != rec.id:
body["existing_batch_id"] = existing_batch_id
log.warning("Duplicate remittance while persisting batch %s: %s", rec.id, exc)
return JSONResponse(status_code=409, content=body)
```
---
## 6. Frontend change
### `src/lib/api.ts`
`ApiError` carries an optional `existingBatchId`:
```typescript
export class ApiError extends Error {
constructor(
public status: number,
message: string,
public existingBatchId: string | null = null,
) {
super(message);
}
}
```
`readErrorBody()` parses JSON and, when status ≥ 400, attempts to extract
`existing_batch_id` and pass it through. `parse837`/`parse835` throw
`new ApiError(res.status, detail, existingBatchId)`.
### `src/pages/Upload.tsx`
New local state:
```typescript
type UploadError = {
kind: "duplicate";
existingBatchId: string | null;
filename: string;
};
const [uploadError, setUploadError] = useState<UploadError | null>(null);
```
On mutation error:
```typescript
} catch (err) {
if (err instanceof ApiError && err.status === 409) {
setUploadError({
kind: "duplicate",
existingBatchId: err.existingBatchId,
filename: file.name,
});
toast.error("Duplicate claim — file not ingested");
} else {
toast.error(err instanceof Error ? err.message : "Failed to parse file");
}
}
```
Inline error panel JSX (renders above streaming results when
`uploadError` is set):
```tsx
{uploadError ? (
<div className="error-panel">
<Badge variant="destructive">409</Badge>
<div className="error-title">Duplicate claim — file not ingested</div>
<div className="error-detail">
{filename} collides with an existing record.
{existingBatchId
? " Open the existing batch to compare, or pick a different file."
: " Pick a different file."}
</div>
<div className="error-actions">
{existingBatchId ? (
<Button onClick={() => navigate(`/batches/${existingBatchId}`)}>
Open existing batch →
</Button>
) : null}
<Button variant="ghost" onClick={() => { pickFile(null); }}>
Pick a different file
</Button>
</div>
</div>
) : null}
```
Styling matches the existing paper-toned surface (slate/parchment) per
the hybrid dark/paper treatment used elsewhere in the app.
---
## 7. Files changed
* `backend/src/cyclone/migrations/0013_drop_claims_unique_constraint.sql` — new
* `backend/src/cyclone/store.py` — `find_existing_batch_for_claim`,
`find_existing_batch_for_remit`
* `backend/src/cyclone/api.py` — both 409 handlers add
`existing_batch_id` lookup
* `src/lib/api.ts` — `ApiError.existingBatchId`
* `src/pages/Upload.tsx` — error state + inline panel
* `src/pages/Upload.test.tsx` — new test file
* `backend/tests/test_api_parse_persists.py` — new test cases
* `backend/tests/test_db_migrate.py` — new migration test
* `backend/tests/test_store.py` — new helper tests
No new dependencies. No config changes.
---
## 8. Out of scope
* A "Replace existing batch" destructive action — requires
`DELETE /api/batches/{id}` and reconciliation cascade handling; deferred.
* General 4xx error UI for other statuses (empty file, parse error,
validation errors). Those already surface in toasts and the streaming
view; a future SP can generalize the inline panel pattern.
* Renaming `patient_control_number` to `subscriber_member_id` to match
what it actually stores. Out of scope for this fix; tracked separately
if it becomes a source of confusion.
---
## 9. Risk
* **Migration reversibility**: SQLite has no `DROP CONSTRAINT`; the
recreation is destructive to schema (but not to data — `INSERT INTO
claims_new SELECT * FROM claims` preserves every row). If the
migration fails mid-way, the implicit transaction (`engine.begin()`
in `db_migrate.py`) rolls back. `PRAGMA defer_foreign_keys = ON` is
required because SQLite otherwise can't drop a table that other
tables reference (`remittances.claim_id`, `matches.claim_id`,
`line_reconciliations.claim_id`, `activity_events.claim_id`). The
deferred checks fire at commit and validate against the renamed
`claims` table.
* **Loss of uniqueness**: after the migration, two claims in one batch
*can* share a `patient_control_number`. This is the intended behavior.
Claim identity is still unique via `claims.id` (CLM01, PK) and the
existing dedup check in `store.add` (`s.get(Claim, claim.claim_id)`).
* **`existing_batch_id` race**: the helper runs after the failed
`store.add` transaction has rolled back. Between rollback and helper
call, another writer could delete the colliding claim. Result: 409
fires with no `existing_batch_id`; UI shows panel without link.
Acceptable — the panel still explains the collision and offers
"Pick a different file."
---
## 10. Test plan
| Test | Asserts |
|---|---|
| `test_db_migrate.py::test_drop_claims_unique_constraint` | After running migrations from v12, `user_version=13`, no `*claims*unique*` index exists, two rows with same `(batch_id, patient_control_number)` insert cleanly. |
| `test_db_migrate.py::test_migration_idempotent` | Running migrations on a v13 DB is a no-op. |
| `test_store.py::test_find_existing_batch_for_claim` | Helper returns None for unknown, returns batch_id for known, deterministic on duplicates. |
| `test_api_parse_persists.py::test_409_response_includes_existing_batch_id` | Upload duplicate; assert body has `existing_batch_id` pointing to the right batch. |
| `test_api_parse_persists.py::test_multi_claim_batch_with_duplicate_member_id_succeeds` | Upload a file with duplicate `member_id` claims; assert 200, all claims persisted. |
| `Upload.test.tsx::test_error_panel_renders_on_409_with_link` | Mock `useParse` to throw `ApiError(409, ..., existingBatchId)`; assert panel visible, link present. |
| `Upload.test.tsx::test_error_panel_renders_on_409_without_link` | Mock 409 with `existingBatchId: null`; assert panel visible, no link. |
| `Upload.test.tsx::test_no_panel_on_non_409` | Mock 400; assert panel absent, only toast. |
| `Upload.test.tsx::test_pick_different_clears_error` | Click button; assert `pickFile(null)` called and `errorState` cleared. |