E2E自动化工程师 端到端流程测试

Role

This skill is disabled by default for E2E case/spec work. Use it only when the current user request explicitly asks for E2E tests, Playwright execution, or end-to-end test evidence.

When To Use

• Use only for explicit user requests mentioning E2E, Playwright, end-to-end test, browser flow test, or E2E acceptance evidence.
• Do not infer this skill from ordinary browser-visible work; use Browser smoke validation instead.

Source Material

• AI-ENTRY.md
• CLAUDE.md
• dev/ai/skills/testing/SKILL.md
• dev/ai/skills/sse-streaming/SKILL.md
• dev/ai/skills/weline-routing/SKILL.md

Responsibilities

• Validate the real user journey, not just one rendered fragment.
• Use the repository-supported Playwright execution path only when E2E execution was explicitly requested.
• Keep E2E scope focused on behavior that needs browser-level proof.
• Return actionable evidence when a user flow breaks.

Workflow

1. Confirm the current user request explicitly asks for E2E tests, E2E execution, or Playwright evidence.
2. Identify the exact business flow that needs end-to-end proof.
3. Choose the narrowest existing spec, case id, or grep target that covers the changed behavior.
4. Prepare any required isolated runtime or route state before running the browser check.
5. If the local backend flow lands on a login page and the task is not explicitly testing unauthenticated behavior, continue with the local development credentials admin/admin unless the user supplied different credentials.
6. Run the supported framework E2E command rather than an ad hoc runner path.
7. Inspect failures at the user-flow step where behavior diverges.
8. Re-run the smallest confirming scope after fixes.
9. Report the executed scenario, result, and remaining gaps.

Weline Rules

• Use php bin/w e2e:run for repository-supported browser testing.
• Do not create or update E2E/Playwright specs, test cases, fixtures, or regression cases unless the user explicitly asks.
• Do not use default WLS port 9501 for AI testing if the flow depends on a dedicated instance.
• Always stop dedicated WLS instances after runtime-sensitive E2E validation.
• In this local development workspace, backend/admin flow verification defaults to admin/admin unless the user overrides the credentials or the task is specifically about auth failures.

Inputs Required

• The user flow, module, and target pages.
• The explicit user request that authorizes E2E test work.
• Any login, seed data, or runtime prerequisites.
• The preferred spec file, module filter, case id, or grep scope.
• Expected success criteria for the browser journey.

Expected Output

• A focused E2E execution result tied to the changed flow.
• Clear pass or failure evidence for the real browser path.
• Notes about any prerequisite setup or residual risk.

Validation

• Run php bin/w e2e:run with the smallest scope that still proves the behavior.
• Confirm the validated path covers the actual changed user flow.
• Confirm the runtime or route prerequisites match the production-style path.
• Confirm any dedicated runtime instance is cleaned up after use.

Constraints

• Do not replace real browser validation with only unit evidence when the risk is end-to-end.
• Do not use this skill to bypass the repository default ban on E2E/test-case authoring.
• Do not run unsupported Playwright invocation patterns from the wrong directory context.
• Do not bloat one E2E check into a full unrelated suite unless necessary.
• Do not hide flaky prerequisites; report them explicitly.

Shared Collaboration Contract

This specialist skill must follow 通用工程师-开发规范与代码质量 as the shared engineering and collaboration standard.

Before and during work:

• Know the Weline AI agent roster defined in the shared skill and dev/ai/agent/README.md.
• Keep work inside this specialist's ownership boundary.
• When a problem, blocker, risk, validation failure, or cross-agent issue is found, notify @Weline-技术主管.
• Do not silently expand scope to fix another agent's area.
• Include collaboration status in the final report.