yunxiafei/openclaw-mission-control
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

chore(e2e): remove gap analysis markdown doc

Browse Source
This commit is contained in:
Abhimanyu Saharan
2026-02-14 16:01:45 +00:00
committed by Abhimanyu Saharan
parent c26dddbce8
commit d1b08b4777
1 changed files with 0 additions and 64 deletions
Download Patch File Download Diff File Expand all files Collapse all files

64
frontend/cypress/E2E_GAP_ANALYSIS.md
View File

@@ -1,64 +0,0 @@
# Frontend E2E gap analysis (Cypress)
This document tracks **critical user journeys** and the current Cypress E2E coverage.
Goals:
- Improve release confidence for high-value workflows.
- Keep tests **CI-deterministic** (flake-averse), favoring **API stubs** over live backend state.
Non-goals:
- Broad UI pixel validation.
- Full end-to-end (DB-backed) integration coverage for every page.
## Current coverage inventory (existing specs)
Located under `frontend/cypress/e2e/`:
- `clerk_login.cy.ts` — smoke sign-in via Clerk testing commands + reach a protected route.
- `organizations.cy.ts` — signed-out redirect + role-based invite permissions UI.
- `activity_smoke.cy.ts` — signed-out redirect for `/activity`.
- `activity_feed.cy.ts` — `/activity` happy/empty/error states; stubs SSE + bootstraps via `cy.intercept`.
## Coverage plan (critical flows)
Legend:
- **Priority**: P0 (must-have for release confidence), P1 (next most valuable)
- **Backend mode**: **Stubbed** = deterministic `cy.intercept` responses; **Live** = depends on real backend state
- **Runtime impact**: rough incremental CI time for the spec (order-of-magnitude)
| Flow | Priority | Coverage status | Backend mode | Expected runtime impact |
|---|---:|---|---|---:|
| Boards list renders + Create CTA for admin | P0 | ✅ Implemented (`boards_list.cy.ts`) | Stubbed | ~10–25s |
| Global approvals: render pending approval + approve/reject | P0 | ✅ Implemented (`global_approvals.cy.ts`) | Stubbed | ~15–35s |
| Skill packs: sync a pack + surface warnings | P1 | ✅ Implemented (`skill_packs_sync.cy.ts`) | Stubbed | ~15–35s |
| Board open + task status move (drag/drop) | P0 | ⏳ Planned | Mostly stubbed (may require DOM/drag stability work) | ~30–60s |
| Task CRUD (create/edit/delete) + error state | P0 | ⏳ Planned | Stubbed | ~30–60s |
| Skills marketplace: list + filters/pagination | P1 | ⏳ Planned | Stubbed | ~20–45s |
| Skills marketplace: install/uninstall dialog wiring | P1 | ⏳ Planned | Stubbed | ~25–50s |
| Gateways list/create (admin gating) | P1 | ⏳ Planned | Stubbed | ~20–45s |
Notes:
- All implemented flows still rely on **auth bootstrapping** (Clerk sign-in) before hitting protected routes.
- We intentionally keep the backend interactions **stubbed** for determinism and speed.
## CI determinism / flake-avoidance notes
Patterns used in these specs (and recommended for new ones):
- **Stub `/healthz`** and **org membership** (`/api/v1/organizations/me/member`) so sidebar/admin gating is stable.
- Prefer **API stubs** (`cy.intercept`) over seeding DB state.
- Stub or neutralize **SSE streams** (where applicable) to avoid race-y updates during assertions.
- Set a higher `defaultCommandTimeout` for auth flows (Clerk helpers can be slow in CI).
- Avoid fixed sleeps; prefer `cy.wait(@alias)` or UI assertions that naturally wait.
- Ignore known non-deterministic hydration errors on auth routes (`Hydration failed`).
## How to run locally
From repo root:
```bash
npm -C frontend ci
npm -C frontend run e2e
```
To run a single spec:
```bash
npm -C frontend run e2e -- --spec "cypress/e2e/boards_list.cy.ts"
```