AgentSkillsCN

full-system-verification

统筹执行全方位的系统测试，涵盖单元测试、集成测试以及边缘场景测试。

SKILL.md
--- frontmatter
name: full-system-verification
description: Orchestrates comprehensive system testing including Unit, Integration, and Edge cases.
version: 1.1

🎯 Triggers

  • •When the user asks to "verify the system", "run tests", or "check for regressions".
  • •During the Verify phase of the PSB protocol (Plan-Setup-Build-Verify-Report).
  • •After significant refactoring or feature implementation to ensure stability.
  • •When core-engineering requests a quality gate check.

🧠 Role & Context

You are the QA Orchestrator. Your responsibility is to ensure the code works as expected under various conditions. You interpret results and decide if the system is "Green" (Go) or "Red" (Stop). You use the evolved verify_system.py runner for robust execution.

✅ Standards & Rules

  • •Quick Check First: Use mode=quick for fast feedback (auto-discovers unit tests).
  • •Interactive: The runner now supports streaming output, so you can see progress in real-time.
  • •Timeouts: Default timeout is 300s (5m), full mode is 600s (10m).
  • •Arguments: You can pass raw pytest arguments in specific mode or appended to other modes.

🚀 Workflow

3. Verification Modes

  • •quick (Default): Auto-discovers and runs core sanity tests.
  • •unit: Runs all unit tests.
  • •integration: Runs integration tests.
  • •edge: Runs edge/stress/security tests.
  • •full: Runs EVERYTHING + Coverage Report.
  • •specific: Runs specific tests passed as extra arguments.

Reports: Test logs are automatically saved to tests/temp/reports/ with timestamp.

  1. •

    Execute:

    bash
    # Standard
python .agent/skills/full-system-verification/scripts/verify_system.py quick

# With Filter (e.g., only login tests in integration)
python .agent/skills/full-system-verification/scripts/verify_system.py integration -k login

# Specific File
python .agent/skills/full-system-verification/scripts/verify_system.py specific tests/unit/test_auth.py
  2. •

    Analyze:

    • •If PASS: Report success.
    • •If FAIL: Analyze the streaming logs (highlighted in RED).
      • •Distinguish between AssertionErrors (Logic) vs ImportErrors (Environment).

💡 Examples

User Input: "I only changed the auth service, verify it."

Ideal Agent Response: "Running targeted verification for Auth Service..."

bash
python .agent/skills/full-system-verification/scripts/verify_system.py specific tests/unit/services/test_auth_service.py

User Input: "Do a full regression."

Ideal Agent Response: "Initiating Full System Verification..."

bash
python .agent/skills/full-system-verification/scripts/verify_system.py full