# TestSprite MCP report — wa_support_node_v1

**Date:** 2026-04-02  
**Source:** Cloud run via `testsprite_generate_code_and_execute` / `node scripts/run-testsprite-execute.mjs`. Detailed machine output: `testsprite_tests/tmp/raw_report.md` (gitignored).

## 1️⃣ Document Metadata

| Field | Value |
| --- | --- |
| Project | wa_support_node_v1 |
| Frontend | http://localhost:3000 (Docker `frontend`) |
| API | http://localhost:3001 |
| Login (seed) | admin@example.com / Admin123! |

## 2️⃣ Requirement validation summary

**Approx. pass rate (from raw report metrics):** 79.31% (23 passed / 6 failed of 29 cases).

| Area | Result | Notes |
| --- | --- | --- |
| Auth / dashboard / tickets / channels | Mostly pass | Core navigation and login flows exercised |
| TC008 Invalid date range on Tickets | Fail | No date-range controls in UI; test plan assumes controls that are not implemented |
| TC011 Channel AI / templates save | Fail | Execution timed out |
| TC013 Invalid empty template | Fail | Save proceeded without visible validation when required textarea cleared |
| TC022 Admin change user role | Fail | Role UI did not reflect change to Admin in agent observation |
| TC025 Audit log actor filter | Fail | No actor filter control on page |
| TC029 Non-admin blocked from audit | Fail | Test used invalid non-admin credentials; could not verify |

## 3️⃣ Coverage and matching metrics

- High-priority frontend plan cases through admin/audit were executed in production Docker mode.
- Several failures are **spec vs product** (missing filters) or **test data** (placeholder non-admin user), not necessarily regressions.

## 4️⃣ Key gaps / risks

1. **Local CLI noise:** Runs logged `checkPortListening tcp timeout: 3000 localhost` and tunnel `503` retries; cloud run still produced results. If runs abort locally, confirm `docker compose` is up and retry; consider TestSprite support if tunnel stays flaky.
2. **Test plan hygiene:** Replace placeholder emails/passwords in generated plans with seeded users (e.g. non-admin) for access-control cases.
3. **Product gaps:** Optional date-range ticket filters, audit actor filter, and stricter channel template validation are candidates for backlog if desired.

## 5️⃣ Plan / PRD maintenance (2026-04-02)

- **`testsprite_frontend_test_plan.json`** extended with **TC030–TC033** (contact detail, company detail + tickets column, link contact to company, deep link contact from company).
- **`standard_prd.json`** updated with routes `/contacts/:id`, `/companies`, `/companies/:id`, Companies / portal feature blurbs, and `code_summary.features` for contacts detail and companies.
- **Canonical PRD:** root **`PRD_v2_WhatsApp_Ticketing_System.md`** bumped to **v2.2** (contacts detail, company link API, companies UI, schema notes). Re-run TestSprite **`testsprite_generate_code_summary`** when you want MCP tunnel runs aligned to the latest file tree.

---

Dashboard links for recordings are in `testsprite_tests/tmp/raw_report.md` (per-test URLs).