# Track 9 Acceptance (Latest QA Run)

## Environment

- Date: `2026-03-05`
- Host: local QA lab (`127.0.0.1`)
- Gateway URL: `http://127.0.0.1:8080`
- Asterisk VM host: `92.38.48.166`
- Database URL (masked): `sqlite:///.data_local/mvp_cc.db`
- Evidence package: [20260305_160002](/e:/Zhan/docs/acceptance/track9/20260305_160002)

## Executed scenario

1. Softphone `1001` called `7000`.
2. Bridge reported AMI connected.
3. `call.started` received from Asterisk source.
4. `call.ended` received.
5. `recording.ready` processed and recording uploaded.
6. Recording playback validated in supervisor shell.

## Validation commands

```powershell
python scripts\track9_preflight.py --base-url <gateway> --check-sftp
python scripts\asterisk_lab_smoke.py --base-url <gateway> --database-url <db> --require-recording
python scripts\track9_check.py --base-url <gateway> --database-url <db> --require-recording
python scripts\track9_collect_evidence.py --base-url <gateway> --database-url <db> --run-checks --require-recording
```

## Result summary

- AMI connected: `yes`
- call.started present: `yes`
- call.ended present: `yes`
- call links present: `yes`
- recording imported: `yes`
- failed bridge events: `0`
- smoke: `PASS`
- track9_check: `PASS`
- preflight: `PASS`

## Risks / open items

- Asterisk dialplan still emits some duplicate/late lifecycle signals in lab runs; bridge reconciliation masks this and keeps platform data consistent.
- For production cutover, keep strict auth and rollback window from Track 9.2 sheet.

## Decision

- [x] GO
- [ ] NO-GO