docs: add MCP-driven execution method to site UAT playbook (GRO-1502) #10
@@ -7,6 +7,23 @@
|
||||
|
||||
---
|
||||
|
||||
## Execution Method
|
||||
|
||||
All UAT is executed by **Shedward Scissorhands** via the **groombook-playwright MCP server**. No manual browser checks or scripted test suites are used for UAT.
|
||||
|
||||
Shedward uses the `mcp__playwright-groombook__*` tool family to navigate, interact with, and verify the site. Each test case step is executed via MCP browser tools:
|
||||
|
||||
- **"Navigate to [URL]"** → `browser_navigate`
|
||||
- **"Locate/Read [element]"** → `browser_snapshot` to inspect the accessibility tree
|
||||
- **"Click [element]"** → `browser_click` with element ref from snapshot
|
||||
- **"Verify [content]"** → `browser_snapshot` and check text matches
|
||||
- **"Check mobile responsiveness"** → `browser_resize` + `browser_snapshot`
|
||||
- **"Check console errors"** → `browser_console_messages`
|
||||
|
||||
Shedward reads this playbook, executes each test case via MCP tools, captures evidence (snapshots/screenshots), and reports pass/fail per test case.
|
||||
|
||||
---
|
||||
|
||||
## 1. Homepage Messaging Path C Refresh
|
||||
|
||||
### Purpose
|
||||
|
||||
Reference in New Issue
Block a user